Kura Technologies is an augmented reality company focused on building the best generation AR glasses, with a 150-degree field of view, 95% transparency, 8K resolution per eye, compact form factor, and high brightness for outdoor

Kura Technologies

“Everything is related to everything else.” Speaking about any ecosystem, we speak about a complex system of interconnected components and the environment in which they exist and with which they interact. All components are linked together through energy flow, some cycles (for example, nutrient cycles in biology), and their environment. The point of interconnection between all elements and the environment is of the highest importance in an ecosystem and this is the thing that distinguishes a from an . system ecosystem A system itself forms a complex and unified whole, while an ecosystem is closely linked to its environment. We can use the term IoT ecosystem instead of IoT system because IoT devices have no value without the environment they exist in. . The main benefit IoT devices bring to people is the data This data is related to the environmental conditions or external phenomena as well as to something within the system. Apart from the relationship with the environment, all devices are connected with each other. The final destination of the data is people who use it.
‍ These three facts bring us to the definition of - a network of interconnected devices that exists in some environment gathers the data and delivers it to people who process and analyze it using modern technologies to achieve a definite goal like building a smart home. an Internet of Things ecosystem Since different groups of people build different IoT applications for their purposes, IoT software development creates numerous IoT ecosystems. These ecosystems can be a simple network with 20 connected devices like smart home or a multi-level structure with a complicated and broad network of devices that requires a sophisticated platform to manage all layers. If we decompose the most complex IoT ecosystem with middle layer into its building blocks, we get the next scheme: s gather the data and transfer it through the to a connected to the Internet that compresses the information and passes it to for further analyses to be later displayed in the to provide with meaningful insights. IoT device securely network gateway the cloud application users ‍ : Thus, we listed 7 components of an IoT ecosystem IoT devices Security Network Gateway Cloud Application Users ‍ ‍These components are sufficient to describe the most complex IoT ecosystem, although most ecosystems don’t need them all. The next step that is expected to be taken is to dive deep into the meanings of these components and describe them in detail. However, we have a better idea. For those readers who want to get a common understanding of these components, we prepare a shortlist with a brief description of them. Those who are looking for detailed information can move to the second chapter where we reveal technical details and additional insights about some of the components. In the final section, we prepared a for you! Bonus abstract Whatever intention you have, Enjoy reading! 7 Components of an IoT Ecosystem Chart Components Details IoT devices Physical devices that interact with the environment. There are two types of them: - are devices that are supposed to gather information about the environment and measure its physical parameters like temperature, motion, people flow, etc. In other words, sensors convert physical phenomena into a digital form. - are devices that perform a physical action on things after they get such a command. actuators turn on the lights when sensors detect a movement within its operating radius. Important notes: as we have mentioned, the main benefit the Internet of Things provides is the data. That means that sensors must transfer the information (about detected movement in our case) to the cloud through the Internet. So when you enter the house and the lights turn on, you most likely don’t become a participant of the Internet of Things ecosystem since the information about you entering the house is not transferred further to the cloud. Sensors Actuators An example: Network is responsible for the communication within an IoT ecosystem between smart things, gateway, and the cloud. : a smart fitness bracelet tracks your heart rate and waits until you connect your smartphone with to transfer the data. After your smartphone gets your heart rate, it stores the data in some installed health apps. The app then will sync with the cloud by transferring your heart rate to cloud servers. The network makes sure that the data will not be corrupted during the transfer via Bluetooth and through the Internet. The network An example Bluetooth through the Internet Security is responsible for access control to the IoT network, the security of data transfers, data leakage prevention, and scanning for malicious software. The security component is presented by firmware and software from security providers, such as Azure Sphere. : many IoT devices have non-existent or very simple passwords for authentication. That led to the emergence of botnets that are also known as “zombie armies”. An army is a group of hacked devices that are connected to the Internet and participate in DDoS attacks on websites. The security component An example Gateway is a physical device that allows for data streams from sensors to the cloud and in the opposite direction. It also performs data preprocessing before the information will be transferred to the cloud. A gateway is not a necessary element since IoT devices can set connections to the Internet by themselves without a gateway as an intermediary. : You probably have a router at your home via which you connect to the Internet. You can consider the gateway as some sort of router (it even may look like a router as a fanless plastic box with buttons), but a bit complicated since it integrates data from thousands of devices, preprocesses it, and has some more responsibilities. A Gateway An example The Cloud is a cloud-based computer resource that is responsible for data storage, deep analysis, and management. In other words, it is a group of computers people get access to through the internet to use their compute capacity for some purpose. The cloud is enhanced by powerful analytic and visualization tools, Big Data algorithms, and Machine learning technology. : imagine you building the Internet of Things ecosystem with 10 000 connected devices that measure the physical parameters of the environment on your field crops. They collect the raw data and send it to your home computer to store it. The computer would get terabytes of information every day. Is its storage capacity enough? Probably not. Furthermore, the raw information is useless until you have an army of humans who can analyze it. Or you can use cloud technologies and get as much storage capacity as you need as well as get all the necessary tools to process and analyze the information. An example of the cloud in the Internet of Things is Cloud IOT core from google. The cloud An example Application s the graphical user interface that provides remote control and management of devices connected to the Internet of Things ecosystem. : you left your smart home and you have doubts whether you locked the door. It’s not a problem! Open the application on your phone and check. If the door is not locked, one click in the app is enough to lock the door. An Application i An example Users are all the people who affect the Internet of Things ecosystem and use it for their purposes. Users comprise people with personal IoT gadgets, researchers who use analytics from the IoT cloud, staff who use the Internet of Things in their operational processes, stakeholders who reap the benefits from huge industrial IoT solutions. IoT ecosystems are supposed to serve people's needs, boost efficiency, improve the standards of living, and quality of life. Users are those who declare business goals and vital postulates that will become the basis for the Internet of Things ecosystem. : You are with a high probability. If there is a fitness bracelet/smartwatch on your wrist, you are a user in the IoT ecosystem. Users Example That’s all you need about the Internet of Things ecosystem. If the information above is sufficient for you, jump to the bottom of the article to from our team! Otherwise, keep reading since we have more information on the components see the Bonus 7 Components of an IoT Ecosystem: Detailed Before we start, it’s worth mentioning that there is no single, officially recognized structure of the Internet of Things ecosystem. Some components are mandatory like IoT devices, some are optional like gateways. The more components the ecosystem includes the more focus it has on finer aspects of the Internet of Things. #1 Component: IoT devices are the layer of sensors, actuators, and smart objects that gather information about the environment and measure physical parameters. ‍IoT devices As we have already mentioned, the basic elements of the Internet of Things ecosystem are sensors and actuators (or just “things”). Sensors are the perception of the IoT system whose main function is to get information from the environment and transform it into data. It’s a rare case when the Internet of Things ecosystem features only one type of sensor or actuator. There are numerous types of sensors where every type has its sub-categories of sensors. We want to mention two of the most common sensors and two of the most essential for improving the ecological state on Earth: - One of the most common and popular. Different industries benefit from these sensors since they can measure the temperature of an industrial machine to check its health, constantly track the temperature of a sick human, or monitor the soil conditions for farmers. Sub-categories: thermocouples, resistor temperature detectors, infrared sensors, etc. Temperature Sensor - A well-known type of device since there are thousands of homes with such types of sensors to save energy of lighting when no one is around. Sub-categories: inductive sensors, photoelectric sensors, ultrasonic sensors. Proximity Sensors - Especially important due to the pollution of the oceans. These sensors can help to monitor the condition of water and detect sources of pollution in real-time. Sub-categories: chlorine residual sensor, turbidity sensor, pH sensor. Water Quality Sensor - They monitor the chemical changes in the air, which is extremely important in big cities where the issues of air pollution continue to worsen. These sensors are also helpful in industrial environmental monitoring, harmful chemical detection, and radioactive detection. Sub-categories: chemical field-effect transistor, hydrogen sulfide sensor, potentiometric sensor. Chemical Sensors #2 Component: Security ‍ - it is the component that encompasses all other components, provides security of data transfer, and prevents unauthorized connections outside the Internet of Things ecosystem. Security In recent years the number of IoT-based DDoS attacks grew dramatically. That is why any IoT system needs a strong security level that protects at least from the most common vulnerabilities. The security level has a wide array of responsibilities, like: . IoT devices may trust the local network to such a level that no further authentication is required. Anyone who connects to the network gains access to all devices within it, making broken authentication issues especially acute. Access control to IoT network while transferring data across the network. The data must be encrypted across the IoT system with protocols, like AES, DES, DSA, and others. Preventing data breaches . In some cases, software bugs can result in the attackers running their own code on the IoT device. It’s necessary to patch software versions when any vulnerability is detected. Scanning for malicious software There are various firmware and embedded security providers to secure the Internet of Things ecosystem, such as Azure Sphere, LynxOS, Mocana, Spartan, Forescout, Symantec, etc. Unfortunately, most Internet of Things providers and IoT device manufacturers overlook even the basic security guidelines. They are: The device boot process must be secured from running inappropriate code pieces. Cryptographic keys must be used to run any command on the devices. It’s especially important in IoT update management. All control commands and information should go through a gateway to avoid direct access to the device outside the network. Whenever a new security breach is detected all IoT devices should install security patches from it. ‍ #3 Component: Network ‍ is the core logistics part of the Internet of Things ecosystem. The network is also called a connectivity layer. It is responsible for all communication within the IoT system: the interconnection of smart things, transferring data and commands between IoT stages, and connection to the cloud. The network There are two ways of communication: takes place locally within a Local Area Network (LAN) between IoT devices and smart gateways through short-range wireless communication protocols. The first way of communication This way of communication is optional because sensors can connect to the cloud directly through the Internet using TCP/IP protocol. However, the connection through non-IP protocols consumes less power since devices connect to local smart gateways instead of trying to access the main server in the cloud. The most popular short-range protocols for the Internet of Things architecture are as follows: WiFi. Bluetooth and Bluetooth Low Energy (or Bluetooth LE for lower-powered devices that generate less data). ZigBee - a universal solutions interlacing all smart devices Near Field Communication (NFC). Radio Frequency Identification (RFID). Sigfox. LoRaWAN. If the system needs to cover long distances in the range of kilometers, it can use Low Power Wide Area Network (or LPWAN) that is designed for wireless data transmission across long distances. occurs when the data is transferred from things to the cloud in cases where there is no smart gateway or in the case of communication between smart gateways and the cloud. The second way of communication The network layer establishes a link between the Local Area Network and the Internet. The basic protocol here is the IPv6 protocol. #4 Component: Gateway is a physical device or virtual platform that serves as an intermediary between IoT devices and the cloud. There are several main functions of the IoT gateways: IoT gateway in the Internet of Things ecosystem. The data flow goes through the gateway from devices to the cloud and in the opposite direction. Control the Data Flow of the transfer of the information in both directions. Ensure the Safety from the cloud to IoT devices. Pass Commands before sending it to the cloud. Gateways filter, aggregate, summarise, and cluster traffic from different devices. Preprocess the Data of the IoT devices since communication through the internet consumes a lot of power, while low energy technologies like Bluetooth low energy don’t. Save the Power of responses to IoT devices. Some devices need a real-time response from the system. Communication with the cloud can take a while, while gateways can be programmed to handle some typical issues and provide instant responses to devices. Reduce Latency #5 Component: The Cloud ‍ is a cloud-based computer resource responsible for data storage, deep analysis, and management. In other words, it is a group of computers people get access to through the internet to use their compute capacity for some purpose. The cloud The cloud is the place where a big pile of raw data from sensors is transformed into neat small piles of valuable information. The most popular vendors of cloud computing are Microsoft Azure and AWS IoT. The cloud can be powered with analytics software, visualization tools, AI, and machine learning for in-depth analyzing and processing of the data. One of the main advantages of the cloud solution is that they are easily scalable which is a vital requirement for building an effective IoT system. #6 Component: Application When software development companies build software for the whole Internet of Things ecosystem they deal with all 7 components and build a system that covers the requirements on all levels. However, IoT application is the tip of the iceberg in IoT software development. is a place where users can interact with the Internet of Things ecosystem. The interaction becomes possible thanks to the graphical user interface where users can look through analytics and reports, control the system, and manage devices.
‍ An application The list of technologies used in the development includes: : C/C++, Python, Ruby, JavaScript Programming languages : Node.Js (Node-Red for rapid prototyping), Ot, IoT.js, Device.js, Eclipse IoT (Kura, SmartHome), AngularJS Development frameworks : Google Assistant, Google Home (Actions on Google), Google Vision, Apple HomeKit, MI Light, Cortana, Alexa Voice Service, Philips Hue, Android Things 3d-party APIs #7 Component: Users ‍ are perhaps the most essential component among all 7 components of the Internet of Things ecosystem. Users have two roles: Users They . The opportunities the Internet of Things ecosystem brings become a basis for valuable insights for all types of users. IoT has a great impact on businesses and the world economy, it changes the existing business models and leads to the emergence of new ones. use an IoT ecosystem for their purposes IoT sensors and applications can become a professional health assistant that measures biometric data of the patient and helps to make a more accurate diagnosis. The Internet of Things transforms all 6 stages in the supply chain management while businesses experience growth thanks to machine downtime reduction and automatization. The impact of the Internet of Things ecosystem is enormous. Users . The Internet of Things ecosystem is supposed to serve people, satisfy their needs, and provide particular information which helps to achieve users’ goals. The IoT ecosystem is built by people, for people, and with a focus on people's needs. define what the IoT ecosystem will do Who can be users: People who use IoT gadgets for personal use Researchers Staff (doctors, warehouse workers, carriers, engineers, etc.) Stakeholders and top managers That is all you need to know about the 7 components of an IoT ecosystem. If you would like to learn more or build the Internet of Things ecosystem for your business, feel free to contact a Sumatosoft team. is a software development company that follows the principle of transparent cooperation and excellence of work. SumatoSoft We're a team of skilled and creative people who specialize in the development of top-quality software for our clients. Bonus! The power of IoT ecosystem or how to cheer up 150 000 people: Source: imgix.ranker Back in 2017, one security expert connected to Internet printers and sent out a message for printing as soon as the printer was turned on. hijacked more than 150 000 The message was a prank and it contained the test: “stack overflow in has returned to his glory, your printer is part of a flaming botnet, the hacker god has returned from the dead. -> YOUR PRINTER HAS BEEN OWNED < - - -” ‍ Under the message, there was an art of a small robot to add extra humor. The prank was supposed to show how certain printers keep ports open by default allowing access to the device through the Internet for everyone. The expert was a good guy and he just wanted to inform users that they need to close this port. He didn’t have evil intentions, so nobody got hurt. However, this case is proof that manufacturers and IoT providers should pay attention to the security of smart devices. Thanks for reading! Co-published here.

Everything You Will Ever Need to Know About the IoT Ecosystem in One Blog Post

Graphs, analytics, and Generative AI. An account of the different ways graphs and AI mingle, plus industry and research news. Is a generative AI preamble necessary for a newsletter focused on , , , and ? Normally, it should not be. However, the influence of generative AI on the items included in this issue was overwhelming. There is a simple explanation for that. Knowledge Graphs Graph Databases Graph Analytics Graph AI It’s been a year since Generative AI burst into the mainstream with the release of ChatGPT. Notwithstanding a rather spotty record both in terms of technical performance and accuracy as well as in terms of business reliability, there’s no denying that Generative AI has captured the attention of executives worldwide. Mentions of “generative AI” on earnings calls have skyrocketed since ChatGPT made its debut, rising from 28 in Q4’22 to 2,081 in Q3’23 — a 74x increase. With the majority of companies being at an early stage in their AI journey, executives feel the pressure to act amid the generative AI rush. What this means is that . Adoption is expected to grow 36% YoY in 2030 according to Forrester, and more than 100M people in the US alone will use generative AI next year. Therefore, vendors are looking to position their products accordingly. generative AI has tremendous mindshare Played right, this can be more than a marketing scheme. There are ways in which graphs and generative AI can complement each other, leading to . responsible enterprise decisions with knowledge-enriched Generative AI Let’s take a brief tour of product offerings and research efforts in that direction unveiled in the last few months. Vectors, graphs, and RAG RAG stands for Retrieval Augmented Generation. It’s a technique through which Large Language Models like ChatGPT can be instructed to use specific background knowledge to contextualize their processing. This offers a conversational interface to proprietary data, making LLMs applicable to business scenarios that require this. RAG is the main reason why the interest in vector databases has skyrocketed in the last year. Like all machine learning models, LLMs work with vectors. Therefore, having a vector database to store information and feed it to LLMs seems like a reasonable choice for RAG. However, it’s not the only one. In fact, as people like Damien Benveniste argue, . Graph Databases may be a better choice for RAG Using graphs, the relationships between the different entities in the text can be extracted and used to construct a knowledge base of the information contained within the text. An LLM is good at extracting that kind of triplet information: [ENTITY A] -> [RELATIONSHIP] -> [ENTITY B] Once the information is parsed, it can be stored in a graph database. The information stored is the knowledge base, not the original text. For information retrieval, the LLM needs to come up with an Entity query related to the question to retrieve the related entities and relationships. The retrieved information is much more concise and to the point than in the case of vector databases. Tony Seale, on his part, argues that vectors need graphs. Embedding vectors are a pivotal tool when using Generative AI. While vectors might initially seem an unlikely partner to graphs, their relationship is more intricate than it first appears. And Ben Lorica makes the case for . boosting LLMs with external knowledge using Knowledge Graphs Graph databases riding the RAG wave RAG is an alternative to fine-tuning LLMs, which seems less demanding and more immediately applicable. What’s more, it can be offered by vendors whose core competency is data management. Since they already have a footprint in organizations, data management vendors can do the heavy lifting of integrating with LLMs. It’s a win-win: executives get to check their gen AI boxes, and vendors get to expand their offering, catching the buzz and keeping clients happy. This is the reason why we’re seeing graph database vendors adding vector capabilities to their products. in its offering in August 2023. The idea is to combine the best of both worlds as part of the . Neo4j was the first one to include vector capabilities roadmap newly appointed Neo4j CPO Sudhir Hasbe announced in July 2023 Neo4j followed that up by . Amazon Neptune followed suit, recently announcing vector search as part of its Neptune Analytics engine; see below. announcing a collaboration with AWS Graph analytics meets HPC Neo4j and Amazon Neptune seem to have been on similar trajectories. Another thing that both vendors unveiled in the last few months is new analytics engines, leveraging parallelism to achieve speedup in processing in graph analytics scenarios previously underserved. As , for Neo4j’s new analytics engine typical examples are analytical queries that traverse most of the graph data. The parallel runtime is specifically designed to address these analytical queries. Hasbe shared The is geared towards three use cases. First, ephemeral analytics. These are workflows where customers just need to spin up a graph really quickly, run some analysis, and turn it off. new Neptune Analytics engine Second, low-latency analytical queries. That involves established machine learning pipelines with feature tables to perform real-time predictions. The third use case is building GenAI applications. Being able to perform vector similarity search when storing embeddings on Neptune analytics means it’s much easier to translate natural language questions into graph queries. To power their new analytics engines, both teams seem to have borrowed a page from high-performance compute (HPC). Neo4j’s implementation was directly inspired by the research paper “Morsel-Driven Parallelism: A NUMA-Aware Query Evaluation Framework for the Many-Core Age.” While we don’t have specific pointers for Amazon Neptune’s implementation, Amazon Neptune General Manager Brad Bebee acknowledged certain similarities. Both teams see a broad subset of graph customers and graph use cases. And both teams have members who are familiar with the literature and techniques coming from (HPC) processing of large-scale graphs. Parallel processing and memory optimization techniques are things that are well-understood in the HPC community. The difference is that HPC researchers are often solving a very specific graph problem on a very specific graph. For a service that has to solve graph problems for many customers, the main challenge is to generalize the techniques that can work well for HPC. Graph database entries and exits Even though Neo4j and Amazon are well-established in the graph database space, there is lots of action in this space beyond these two. Most vendors are working on positioning themselves in the generative AI landscape and/or adding conversational interfaces to their products. But that does not mean that generative AI is all there is out there – there are other use cases as well. , aiming to tackle complex problems at scale. Aerospike Graph is a new entry in the graph database market Aerospike started out as a key-value store. Eventually, the initial offering expanded to include the document model (JSON) as well as a SQL interface via Starburst. Graph was the next step, with adoption being reportedly customer-driven. A team was built for Aerospike Graph, including Apache TinkerPop founder Marko Rodriguez and key contributors to the project. They helped Aerospike create a graph layer that interfaces with the core engine in a way that scales out horizontally in a shared-nothing architecture. Coincidentally, it was only a few days after Aerospike Graph’s official unveiling that another database vendor with a similar profile who had recently entered the graph database market announced its exit. In 2019, , citing similar reasoning: they wanted to offer performance and scalability. In 2023, : Redis introduced its graph database they wound down RedisGraph, saying “Many analyst reports predicted that graph databases would grow exponentially. However, based on our experience, companies often need help to develop software based on graph databases. It requires a lot of new technical skills, such as graph data modeling, query composition, and query optimization. As with any technology, graph databases have their limitations and disadvantages. This learning curve is steep. Proof-of-concepts can take much longer than predicted, and the success rate can be low relative to other database models. For customers and their development teams, this often means frustration. For database vendors like Redis, this means that the total pre-sales (as well as post-sales) investment is very high relative to other database models”. Scoping and building knowledge graphs Building knowledge graphs is supposedly a huge and terrifying project, as . But that perception is mostly coming from software engineers who think that a knowledge graph has to be enormous. Mike Dillinger notes Tech organizations are overflowing with engineering managers who think (or expect) that engineers can do anything and everything. And there are precious few product managers who are familiar enough with knowledge graphs to frame knowledge graph building in terms that managers and engineers can grok and buy into. The end result is that organizations don’t implement crucial but unfamiliar tech, like knowledge graphs. Contrary to widespread assumptions, , creating and curating knowledge graphs is not an inherently manual process, even if quality management requires expert review. He shares some of the latest research efforts for assisted knowledge graph creation. Dillinger adds The topic of assisted knowledge graph creation has received renewed attention in the last few months, again owing to the generative AI momentum. There are good reasons for this. LLMs can assist in , as Rahul Nayak goes to show. converting any text into a graph of concepts One such effort that has drawn lots of attention is . MechGPT first breaks down texts into small chunks. Each chunk is fed to a general-purpose LLM that generates question-answer pairs summarizing key concepts. MechGPT, a language model specifically fine-tuned for constructing knowledge graphs Bonus track 1: Yejin Choi in conversation with Bill Gates, . Α combination of LLMs (magic but opaque/subsymbolic) with knowledge graphs (explicit/symbolic) is the path forward. summarized by Jesús Barrasa Bonus track 2: by Peter Lawrence. Prompting an LLM with an ontology to drive Knowledge Graph extraction from unstructured documents Ontology modeling for Knowledge Graphs with SHACL What is an ontology? That may seem like a trivial question for people familiar with knowledge graphs, but as , it is not. Cagle defines an ontology as a set of schemas that collectively establish the shape of the data held within a named graph. Others may prefer different definitions. Kurt Cagle goes to show that in the world of knowledge graphs, an ontology is a domain model defining classes and properties. Classes are the types of entities (instances) in the graph, and properties are the attributes and relationships between them. Ontologies define the structure of graphs and allow tools to make more sense of them. Holger Knublauch writes Either way, there seems to be convergence around the notion of using SHACL, the Shapes Constraint Language, for Ontology Modeling. Prior to the introduction of SHACL, knowledge graph validation would mostly be manual or reliant on OWL (Web Ontology Language) constraints. OWL constraints, however, are counter-intuitive. Knublauch followed up on his SHACL tutorial with . and . Radostin Nanov has also written a SHACL guide in 3 parts: , , and . And Ivo Velitchkov with Veronika Heimsbakk maintain a . Part 2 on Qualified Cardinality Constraints Part 3 on SPARQL-based Constraints Learning how SHACL can tame unruly data graphs applying SHACL to data and handling the output the internals of a SHACL engine Wiki on SHACL Bonus track 1: . This report presents a curated collection of ontologies that are highly relevant to ICT domains and vertical sectors, considering their maturity, prominence, and suitability for representing linked data in the semantic web. too. Bonus tracks 2 and 3: and a . A Landscape of Ontologies Standards Landscape Ontologies can contribute in Explainable AI SPARQLing anything SPARQL Wiki Applied Graph AI use cases from DeepMind DeepMind has been one of the forerunners of Graph AI. In the last few months, DeepMind has shared more details about their use of Graph AI in a couple of high-impact use cases. In a paper , DeepMind introduces , a state-of-the-art AI model able to make medium-range weather forecasts with unprecedented accuracy. GraphCast predicts weather conditions up to 10 days in advance more accurately and much faster than the industry gold-standard weather simulation system. published in Science GraphCast GraphCast is a weather forecasting system based on machine learning and Graph Neural Networks (GNNs), which are a particularly useful architecture for processing spatially structured data. DeepMind has enabling scientists and forecasters around the world to benefit billions of people in their everyday lives. open-sourced the model code for GraphCast, is a GNN-based system that discovered 2.2M new crystal structures including about 380k stable structures. Novel functional materials enable fundamental breakthroughs across technological applications from clean energy to information processing. Graph neural networks trained at scale can reach unprecedented levels of generalization, improving the efficiency of materials discovery by an order of magnitude. DeepMind GNoME Bonus track: . A conceptual review of recent key advancements in this emerging field, aiming at optimization and machine learning researchers by DeepMind’s Petar Veličković et.al. Combinatorial optimization and reasoning with Graph Neural Networks Graph Levels of Detail Graph-based data is ubiquitous in enterprises across the industry verticals and is increasingly needed for machine learning use cases. Graph technologies are available, though not quite as widely used yet in comparison with relational databases. Even so, interest in knowledge graph practices has grown recently due to AI applications, given the benefits of leveraging graphs and language models together. A frequent concern is that graph data gets represented at a low level, which tends to make queries more complicated and expensive. There are few mechanisms available — aside from visualizations — for understanding knowledge graphs at different levels of detail. That is to say, how can we work with graph data in more abstract, aggregate perspectives? While we can run queries on graph data to compute aggregate measures, we don’t have programmatic means of “zooming out” to consider a large graph the way that one zooms out when using an online map. This leaves enterprise applications, which by definition must contend with the inherently multiscale nature of large-scale systems, at a distinct disadvantage for leveraging AI applications. a survey of related methods to date for level-of-detail abstractions in graphs, along with indications for future work. Paco Nathan presents Graph and Large Language Models research Wrapping up with a collection of research on various aspects of combining graphs with large language models. Paco Nathan also lists a collection of research efforts to combine Graph ML with Language Models . here IEEE researchers present . Their roadmap consists of three general frameworks. KG-enhanced LLMs, LLM-augmented KGs, and synergized LLMs + KGs. Cf. the previous YotG issue . a roadmap for unifying large language models and knowledge graphs here Similarly, researchers from The Hong Kong University of Science and Technology (Guangzhou), The Chinese University of Hong Kong and Tsinghua University present . They propose a taxonomy that organizes existing methods into three categories based on the role (i.e., enhancer, predictor, and alignment component) played by LLMs in graph-related tasks. a survey of graph meets large language model Researchers from ETH Zurich, Cledar, and the Warsaw University of Technology introduce (GoT): a framework that advances prompting capabilities in large language models (LLMs) beyond those offered by paradigms such as Chain-of Thought or Tree of Thoughts (ToT). Tony Seale . Graph of Thoughts summarizes the approach Michael Galkin et.al. present . ULTRA can generalize to new KGs of arbitrary entity and relation vocabularies, which serves as a default solution for any KG reasoning problem. Researchers from Monash and Griffith present , a method that synergizes LLMs with KGs to enable faithful and interpretable reasoning. ULTRA, a single pre-trained reasoning model for Knowledge Graph reasoning Reasoning on Graphs (RoG) A growing number of experts from across the industry, including , , and industry analyst firms like , point to Knowledge Graphs as a means for improving LLM response accuracy.  To evaluate this claim, data.world researchers came up with a that examines the positive effects that a Knowledge Graph can have on LLM response accuracy in the enterprise. academia database companies Gartner new benchmark The benchmark compared LLM-generated answers to answers backed by a Knowledge Graph, via data stored in a SQL database. point towards a significant improvement in the accuracy of responses when backed by a Knowledge Graph, in every tested category. Findings is a scalable, inductive, and interpretable path-based Graph Neural Network on Knowledge Graphs. It can be used to make ChatGPT more factual by equipping it with knowledge graph reasoning tools. It’s open source and integrated with ChatGPT. A*Net Google Research presents . A comprehensive study of encoding graph-structured data as text for consumption by LLMs. Talk Like a Graph: Encoding Graphs for Large Language Models Jure Leskovac et al. introduce an end-to-end deep learning approach to directly learn data spread across multiple tables called . The core idea is to view relational tables as a heterogeneous graph, with a node for each row in each table and edges specified by primary-foreign key relations. Relational Deep Learning Subscribe to the Year of the Graph Newsletter Also published . here

The Year of the Graph Newsletter Vol. 25: The Fusion of Generative AI and Graph Technologies

Computer vision (CV) is a sub-domain of the field of Artificial Intelligence that is aimed at helping machines to identify and understand content in digital video or images. Simply put, CV enabled machines to “see” the world like we humans do and use that knowledge to augment human efforts. This technology uses cameras and computers instead of the human eye to identify, track and measure targets for further image processing. The field of computer vision has progressed tremendously in the last decade thanks to: exponential growth in the ability to complex machine learning or deep learning models through higher computation efficiency in the form of GPUs or purpose-built hardware accelerators, decreasing hardware costs, rapid advancement of new technologies (better algorithms, automation, and easy-to-use tooling), and advancement of camera technology that allows higher quality images to be captured and used for CV tasks. How does Computer Vision work CV tasks can be performed by using either traditional image processing techniques or modern deep learning (DL) based techniques. Traditional image processing techniques used statistical methods and did not involve learning from patterns (i.e., did not need training data). OpenCV is a popular image-processing library and is widely used today to pre-process images. However, the performance of such models has been restricted by the scene, lighting, and illumination. It also doesn’t scale well with new use cases. Deep Learning based techniques offer much higher performance across a wide variety of scenes and lighting conditions but rely on large training data as well as powerful hardware to run the computations. DL-based techniques constitute the state-of-the-art for CV and we will focus on DL-based CV techniques only for the rest of the article. On a very high level, computer vision can be described as a 3-step workflow: Capture an image from a camera or video (video is nothing but a series of images) Analyze the image for the presence of objects or features by passing the image through special deep learning (DL) models If needed, pass original images along with the output of the DL model into another DL model to identify activities and complex higher-order traits In theory, any image from a camera can be used for CV tasks. However, even with the most light-weight algorithms, CV tasks are computationally intensive and require massive data sets to train the models on (it could cost hundreds of thousands of dollars to train a state-of-the-art DL model) so in many cases where real-time results are needed, the image or video’s resolution (thereby size) is reduced along with other pre-processing steps such as contrast enhancement, noise reduction, etc. to speed up computation. Modern CV technologies rely heavily on deep learning models, specifically, deep Convolutional Neural Network (CNN) models, that have been trained on hundreds of millions of images and offer considerably good precision and recall (more on this later in the article). A CNN model helps understand the contents of the image by breaking it down into objects with labels that describe what the object is. The labels used are part of the dataset used to train the model. Today in many tasks, CV algorithms are surpassing human performance. Some state-of-the-art deep learning models for CV are discussed later in the article. Key Computer Vision tasks There are dozens of tasks that can be performed with Computer Vision. Some major tasks are described below, with the most important task being object detection. Object detection includes 3 sub-tasks: Object Detection: a) detecting an arbitrary number of objects in images or videos, locating them within the image with a bounding box, classifying which classes they belong to (humans, animals, vehicles etc.) and labeling them as such. Object detection tasks help answer the question of What type of objects are present and where they are located. Object detection is a super-set of tasks like person detection, vehicle detection, animal detection, plant detection, etc. An object detection model can detect many different object types, but detection performance might be different for different objects based on the data used to train the model. Object detection is the primitive used for many other analytics tasks such as person counting, loitering, measuring queue length, average dwell times, etc. a type of object detection focused on detecting human faces within an image/video Face detection: people present in the image/video are recognized by their names and by matching their faces with a database containing names of people and their faces. Facial recognition: estimate the position and orientation of humans in the image/video. Pose detection involves detecting different “parts” of the human body (such as right elbow, left knees, right shoulders, etc.) and their pose relative to other parts to estimate the overall pose of the human body in order to determine the actions or activities (such as sitting, walking, running, hitting, etc.) that are being done by the humans in the image/video and get more in-depth understanding of the human body language to estimate complex higher order tasks such as determining human behavior, estimating stress, demeanor, mood, etc. Pose estimate is used in the world of online and remote fitness classes to determine if the exercises and activities are being performed in the optimal way to prevent injuries. Pose estimation also has other interesting applications such as training robots, motion tracking in augmented reality and virtual reality, fall detection, etc. Pose estimation: Identification of characters (alphanumeric numbers and symbols) in an image. Automatic License Plate Recognition is an application of optical character recognition techniques Optical character recognition: Key Computer Vision Models to know There are many different CV model architectures with new versions coming out every few months. But each model has its own pros and cons and is good for certain use cases and tradeoffs between speed, detection accuracy, and efficiency. YOLO stands for You Only Look Once is one of the most popular detection algorithms that can detect objects in real-time at the expense of slightly reduced detection accuracy. It is widely used across the industry from large corporations to startups. The first generation of YOLO (Yolo v1) was launched in 2016 and it was the fastest object detector of its time. Since then, there have been multiple iterations with every new release improving the speed, accuracy, and efficiency of object detection. YOLOv7 was released in July 2022 and is the fastest and most accurate object detection model as of writing this article. YOLO also has many optimized versions for special use cases, such as YOLOv7-tiny which is a lightweight version of the main YOLOv7 and optimized for edge AI and efficient yet performant enough to run on edge GPU devices, and YOLOv7-pose that delivers state-of-the-art pose estimation performance. YOLOv7: stands for cascade-mask region-based convolutional neural network and is state-of-the-art in terms of image segmentation. Image segmentation is the task of clustering parts of an image together that belong to the same object class by annotating pixel by pixel. Image segmentation goes deeper than just object detection and gives a richer partitioning of the elements in the image. Mask R-CNN is the state of the art in terms of image segmentation and instance segmentation. Image segmentation has broad applications such as detecting tumors, road sign detection from different angles, autonomous driving assistance systems, precision agriculture, and use for sensing through drones and UAVs. MASK R-CNN is optimized for detection accuracy at the expense of speed of detection. However, YOLOv7 achieves slightly better accuracy (2% higher than Cascade-Mask R-CNN) at a significantly increased detection speed (509% faster). Cascade-Mask R-CNN: is lightweight and designed for mobile CV applications that are restricted by available power, available computing, and hardware to run. MobileNet was a popular model but recent advances with YOLO architecture, especially YOLOv7 significantly improved detection speed even in processing power-constrained environments while still providing good accuracy. MobileNet-SSD V2: detects human body, hand, facial, and foot key points (in total 135 key points) in multi-person images in real-time and is a must-have for pose estimation. Knowing pose and orientation has a lot of applications, especially when tracking the movement of the human body is involved. For example, in fitness, pose estimation is used to identify if the athlete is following the right movements and if the pose will prevent injuries. OpenPose detects multiple people so with decent accuracy and in real-time. OpenPose also enables single-person tracking that can be used to speed up detection and visually smooth images and video by predicting where the movement will be. OpenPose: Edge computing for high-performance CV applications Edge computing is when computing happens near the source of data rather than in data centers optimized for large-scale compute workloads. For example, when you are typing in your mobile phone and it tries to predict the next word, that prediction is happening on the mobile phone itself close to the data (your previous chats or conversation) and is a type of edge computing. Other examples of edge computation devices are laptops, TVs, special edge computing hardware, Raspberry-pi, etc. Edge computing has tremendous value for computer vision due to bandwidth reduction, increased speed, and privacy as well as other security benefits. Edge computing is also more efficient than cloud computing (since data is not passed between the edge device and the cloud). Because computation happens so close to the user, edge computation allows for real-time and ultra-low latency performance. Imagine in a self-driving car example, if the car had to evaluate the scene by sending each frame to the cloud and waiting for analysis, it would never be able to respond in milliseconds to avoid accidents. Also, edge computation is resilient to network downtime and outages (imagine the car going through an area where the cellular network dropped, and not being able to process and stop in time for a pedestrian crossing the street. As a corollary, since data doesn’t go to the cloud for processing, edge computation minimizes internet bandwidth usage and reduces the cost of data transmission back and forth from the cloud servers. Also, because computation is more efficient, edge computing consumes less energy. Lastly, edge computing allows for techniques such as federated learning to happen that give best-of-breed machine learning solutions while protecting privacy. This is because data never leaves the end devices and thereby is less prone to being intercepted or leaked from central servers. This allows for high-accuracy computation to happen on customers’ personal and confidential data to give highly personalized experiences and best-in-class UX without the threats and risks of data being intercepted or breached or hijacked. The disadvantages of edge computing have been the complexity of working with various edge devices and compatibility with the device classes. The distributed nature makes edge computing hard to standardize and requires heavy lifting for a lot of different architectures and systems. However, in this world of diverse classes of devices, there are a few edge devices that stand out in terms of acceleration of CV inference. Note that model training is still ideally done in the cloud and even with the most powerful GPUs, training a modern deep neural network can take anywhere from hours to a few days. NVIDIA Jetson NVIDIA Jetson is a family of edge computers that is optimized for running AI / ML / CV applications on the edge. It is ideal for students, tech enthusiasts, DIYers as well as professionals to quickly build and deploy edge AI applications. It comes with its own SDK and a complete ecosystem, comprehensive tutorials, training modules, and getting-started guides. Due to the strong ecosystem support as well as being high performance and power efficient, it is one of the most popular platforms for edge AI applications today. Additionally, the platform has been designed for flexibility so modularity is built into each component. The SDK comes with all the required libraries, debuggers, and APIs, and supports popular Machine Learning frameworks like Kera, Caffe, and TensorFlow that just makes development on NVIDIA Jetson very easy and fast. There are different modules available in the market, starting with Jetson Nano which is optimized for building and deploying small AI systems and is great for beginners and students. Don’t mistake the nano in its name to mean small - it is power packed to run parallel neural network workloads, processing data from high-resolution sensors and running hefty workloads, and is ideal for building proof of concepts. The TX2 raises the performance bar and is designed for beefy workloads with high efficiency and speed. The module carries NVIDIA Pascal GPUs and is 2.5x more powerful than the Nano, allowing running deep neural networks with higher accuracy. Jetson Xavier is said to be the smallest supercomputer in the world for edge computation workloads and is designed to run parallel neural networks and process data from multiple high-resolution sensors simultaneously. Jetson AGX Xavier is developed with fully autonomous applications in mind and 20x the performance of the TX2 and is the highest-performing module in the Jetson series. Jetson AGX is ideal for autonomous mobility applications, warehouse robotics, and large surveillance drone use cases. Google Coral Google Coral is a Tensor Processing Unit processor-powered edge computing device that is ideal for fast neural network inferencing. Google Coral is basically a purpose-built inference engine for Tensor workloads. Google Coral has 2 components - a compute device that is an edge compute device sized slightly larger than a credit card, and also has a USB accessory that is an AI accelerator. Google Coral is a high-performance, extremely low power-consuming edge computer and has a pretty good cost-per-FPS processing ratio. Intel Neural Compute Stick 2 Intel Neural Compute Stick 2 is Intel’s answer to edge computing for AI workloads and is powered by the Intel Movidius Myriad X VPU (vision processing unit) chip that is optimized for CV applications. Highlights of this device are high performance and ultra-low power and can be used for a wide variety of applications including VR/AR headsets and security solutions. It comes in a tiny, USB thumb drive form factor and is an accelerator for speeding up neural network inference on the edge in a cost-effective way. Essentially, they can be thought of as purpose-built co-processors that take the burden of some special computation from the CPU that is required for CV applications. As with NVIDIA Jetson, Intel Neural Compute Stick 2 comes with its own SDK (called the Myriad Development Kit) that contains the necessary libraries, tools, and APIs to quickly implement solutions. Because of the USB interface, it can be plugged into any device that has USB ports and is easy to get started. There are extensive tutorials and guides and a strong community that has tons of getting-started guides and examples to help anybody get started quickly. The Neural Compute Stick can take in multiple high-resolution video streams at the same time and with high frame rates.

An Intro to Edge Computer Vision: Technologies, Applications, Use Cases and Key Models

According to , only 5% of crypto entrepreneurs are female. Today I decided to interview one of them. Please meet Milana Valmont, founder and CEO of , the first Layer 1 network that enables market access to any digital. Read more about her journey in crypto, an idea of KIRA, DApps market, the current pitfalls, trends, and the future of DApps. Crypto Head research KIRA Please share with us your background or story My background is in finance. I graduated from Fordham University with a degree in economics and psychology and I worked in corporate finance in New York. First as a Private Equity Analyst then as an Investor Relations Manager. I joined the blockchain space in early 2017. I have worked as a project manager for Adcoin.com and served as an adviser for Sentinel dVPN. In 2018 I joined Knoks platform as head of the strategy, and by mid-2019 I dedicated my full attention to KIRA as CEO. When and how did you discover crypto? When did you join this industry? In early 2017 I talked to my friend who was a cyber security expert and we were discussing Bitcoin and Ethereum. He told me that he was trading cryptocurrencies but I didn't have a lot of background in it, so I started to learn more about buying cryptocurrencies and how different technologies worked in blockchain and why they are important. I joined a couple of crypto Slack and telegram groups and became more involved within the community. I started to get more involved in the space when I became a community volunteer for Binance exchange when Binance was first starting out in Tokyo, which was around May 2017. Later that year I quit my corporate job in New York and started focusing more professionally on crypto. What are you working on right now? Right now I'm working on KIRA. I'm the founder and CEO. KIRA is Layer 1 Network that allows users to deposit and stake any assets whether it's any cryptocurrencies or real-world assets issued in the digital form. All tokens staked on KIRA remain liquid and are represented 1 to 1 by transferable derivatives held trustlessly by the stakeholders themselves. Through KIRA's unique offering stakeholders would be able to benefit from the block and fee rewards generated by the network, as well as leverage stake their assets on multiple PoS chains and DeFi platforms simultaneously. To summarize: KIRA  is an L1 network capable of scaling value at stake without limiting the liquidity of users tokens. KIRA rose up out of the idea to connect crypto assets with the real-world economy Why did you decide to build a decentralized network for DApps? How was the idea of KIRA born? Co-founder & CTO Mateusz Grzelak and I were both a part of the Interchain ecosystem for a few years before we started KIRA. We shared a vision that there's not going to be a single masterchain network that scales infinitely and provides all features that you might imagine. I think it was more rational to assume that in the long term it will have thousands of projects interconnected together and just in the world with thousands of smaller companies on the internet offering their services rather than just one large single website like Amazon. So we wanted to utilize Web3 structure and make sure that people utilize 100% of their capital with the minimum possible risk involved without fears of liquidation, fears of market volatility and we didn't want those people to just sit on their asses. KIRA rose up out of the idea to connect crypto assets with the real-world economy so that people can stake whatever assets they have in the digital world and earn rewards in a safe manner. What does Multi-Bonded Proof of Stake (MBPoS) mean? It’s a proprietary consensus that we created. As you probably know Bitcoin uses Proof of Work consensus that allows scaling of security with the growing hash rate but is unable to scale transaction throughput and does not offer finality. You are also likely familiar with Proof of Stake that now a lot of projects like Cosmos and Polkadot use. However, that consensus offers scalability of transactions and finality does not scale the security well. We’ve created a customized consensus which is called Multi-Bonded Proof of Stake (MBPoS) and which allows you to get the best of both worlds by allowing staking of multiple assets to secure the network without locking up the capital The consensus further puts the power directly into the hands of individual network participants who can decide which DApps they want hosting on KIRA thus ensuring their own security while offering maximum possible utility for the end-users. Instead of the network getting secured by just one token which is not secure at all because it centralizes over time in hands of a few as its value fluctuates unpredictably on the market, people can stake stablecoins and even digitized real-world assets such as NFTs representing the real estate. Furthermore, because voting power and stake are NOT correlated with each other then no one can circumvent the consensus by hoarding or stealing digital tokens. Effectively you can quantitatively measure the security of the chain as the value at stake is distributed among a possibly large number of unique (non-Sybil) validators who operate the chain. All this brings clarity to the definition of what a “secure” means in the context of decentralized networks. What’s Initial Validator Offering (IVO)? How does it work? IVO is a crowdfunding platform. It works by enabling any validator or any project to deploy crowdfunding on KIRA where delegators (your investors) can earn that new token or mint the new NFTs without spending or locking up their existing capital. If you want to invest in the project right now there are crowdfunding platforms for Initial DEX Offering (IDO) so you send USDT or Ethereum and in return, you get tokens you invested in. With the KIRA crowdfunding platform, it’s going to be different. The project that is raising money will host an IVO on KIRA by operating or partnering with existing validator nodes and the investors who are interested in acquiring new tokens will delegate their assets to the IVO validator/s. It doesn't matter if it’s USDT, Bitcoin or Ethereum so they delegate the asset and get rewarded with a new token. And if they decide that the project isn’t performing as they expected they can undelegate the asset and they still get to keep the token that they earned as a reward. It reduces the risk for investors to invest in projects as instead of exchanging your valuable asset such as stablecoins, NFTs or Ethereum for the new illiquid token, you can just stake your existing coins without losing custody or liquidity. What problem is KIRA going to solve? The key problem is the lack of security in the context of many small networks existing within the internet of blockchain to offer access to the Web3. The security assumptions of all currently existing Proof of Stake networks only work when the network exists in a vacuum like entirely alone. Cross-chain applications utilizing small PoS chains as their backbone consensus stop being secure when large values of assets are deposited to them. When the value of deposits from outside of the network exceeds the native value at stake the network operators become incentivized to misbehave - fork the chain and steal all deposits. Imagine sending $1B to the small chain with just a few validators and only $10M at stake vouching for the honest operation of the “dApp”, the incentives of $990M pose a huge risk of validators collusion to cheat. Please note that we also have to consider attacks where someone acquires a majority of stake/voting power in such a disadvantaged chain - as the costs of performing the attack (slashing) become justifiable. One of many examples of such exploits is . centralized exchanges voting on governance proposals to help wealthy individuals acquire a “decentralized” platform The key problem is the lack of security… KIRA solves such security issues because there is no cap to the value at stake that can vouch for the chain security as well as there is no correlation between voting power and the wealth of any individual in terms of the native/governance token. No one can steal, hoard, borrow the coins or collude with just a few individuals to circumvent the consensus and everyone can transparently measure if the value at stake is sufficient for their own deposit to be safe. In KIRA there is no limit in terms of how many validators can be part of the network. However, each one of them must be an individual (non-Sybil) as the voting power is split equally between each one of them. The social consensus of validators is used to ensure that each of the individuals operating the chain is indeed an individual. It's a totally different approach to consensus. We're not utilizing artificial token to secure our network and plutocratic government principles which makes the issue of security even more prominent. KIRA uses real-world and any type of digital assets to secure itself, allowing stakeholders to benefit from the transaction fees and block rewards. All the accumulated value vouches for the honest operation of the chain and deployed their dApps. Who are the key users of your solution? The key users of our solution are those who are interested in generating passive income from staking and operating nodes on the network. The idea of staking is pretty big and it's not just staking, farming, but also staking any assets that you wish to use. There is a lot of hype around NFTs. What can you do with NFTs rather than buying them for speculative purposes and hoping that they will increase in value? If you have any NFT collection you can stake them on KIRA and earn rewards on top of the floor price increase. Of course under the condition that the network governance whitelists the NFT collection on KIRA because it adds a lot of community (fee revenue) or security benefits to the chain. So basically it’s retail and anyone who is interested in staking or running validator nodes. As soon as KIRA L1 is live our focus will be on enabling L2 dApp hosting to attract retail users, that is existing DeFi  & CeFi consumers. When our native crowdfunding system is launched it's going to be not just retail who want to earn benefits and rewards from staking, but also investors who are interested in investing in other projects through staking and the projects themselves which are interested in using our crowdfunding platform to raise capital. What’s the current situation with DApps market? Any pitfalls? Yeah, there are a lot of pitfalls. I think the main problem with DApps is the lack of trust and quality. DApps running on the Proof of Stake are often secured by a small validator set or one single governance token and that creates a huge security threat. Anyone can launch their code and very rarely any of it is vetted by any third parties. All this creates potential issues not only for the end-users but also for the network operators themselves that might find themselves in the jurisdiction that might suddenly prohibit the operation of certain applications such as DEX’es without KYC in the USA. KIRA tackled all the mentioned pitfalls and is designed with the future-proofing of the Web3 ecosystem in mind. The main problem with DApps is the lack of trust and qualit From the social point of view, I see a huge hype around NFTs. Eventually, projects like Ethereum utilized the usage of NFTs as a speculative value. I think a lot of projects become redundant as well. What trends will we see in the near future? In terms of the trends people who will go towards projects that are: first of all, quantitatively trustable and secure which involves securing the network and decentralization of the governance allow you to have custody over your assets and decisions, you don't have to go through million different steps and risk your capital in order to generate passive income and are legally compliant, but not in the way that is disinter changed to the user. What’s the next stage of DApps development? The Layer 2 scaling solutions will definitely bring the breath of the fresh air, responsiveness, low fees, accessibility, and user-friendliness to the crypto ecosystem. All that while eliminating a centralized middle-man and the gatekeepers deciding who can access which financial products and who can not. I also see the tighter global regulations coming for the network operators and even the developers themselves as the world slowly becomes to normalize crypto as an intangible part of our daily lives.

The Usability of DApps Should be as Simple as Venmo: Interview with Milana Valmont, CEO of KIRA

At Koyeb, our mission is to provide the fastest way to deploy applications globally. We are building a platform allowing developers and businesses to easily run applications, a platform where you don't need to think and deal with the resiliency and scalability of your servers: . a serverless platform Ironically, a serverless platform is actually full of servers. As a cloud service provider, we operate the infrastructure for you and abstract it as much as possible. The initial technological choices you make as a cloud service provider are crucial. Not only do they determine the features you can develop, but also the resiliency and scalability of your entire platform. We decided to build our own serverless engine, one that would not be limited by existing implementations. The first version of Koyeb was built on top of Kubernetes and allowed us to quickly build a working cloud platform. After a few months of operating with this version, we decided to move user workloads from Kubernetes to a custom stack based on Nomad, Firecracker, and Kuma. We wrote this blog post to explain why we moved from Kubernetes to a custom stack to power our users' serverless workloads. This post is written from our point of view as a cloud service provider. Koyeb's requirements: Fast, Global, Secure, Scalable Building on top of the right abstraction layer The limits of Kubernetes The stack behind Koyeb Serverless Engine: Nomad, Firecracker, Kuma on BareMetal... and Kubernetes Koyeb's requirements: Fast, Global, Secure, Scalable Before diving into the technicals, let me outline the requirements which led us to where we are today. At Koyeb, we are building a next-generation serverless platform that should be the fastest way to deploy applications globally. There are four core pillars we want to achieve: : We want the platform to be quick, both when you deploy your application and when the application is running. We believe deploying a software update should take minutes or ideally seconds. No developer wants to wait dozens of minutes before an upgrade is deployed, especially when working on development or pre-production environments. Execution should also be fast and your workloads should not be slower in the cloud than on your laptop. Fast : For businesses, deploying globally is rarely optional in 2021, and it is definitely not an option for us as we have customers all around the world. Whether you're operating a global B2C brand or a B2B company, if you have users all around the world, you need a simple and effective way to operate in multiple regions. Compliance and performance are the two drivers here. Global : As a cloud service provider, security is crucial, especially as we operate a multi-tenant environment. Isolation between customers should be reliable, both at the host and network level. From a user perspective, security should be built-in and should not be an option. Secure : To be competitive on the market, we need to offer a scalable platform for our end-users and we need to build a platform able to support both your growth and our own growth. We've already experienced the pain of struggling with growth . To better support our customers all around the world, we want to be able to run on different cloud service providers depending on locations and to respect sovereignty requirements. Scalable in a previous life Now that we have a list of requirements, let's move to the technical options. Building on top of the right abstraction layer The options We considered all levels of abstractions to build a full-featured platform that would be able to support our mission and requirements. We had roughly 4 options, below is the list sorted from the highest level of abstraction to the one where you deal directly with hardware: : Building on top high-level function and container primitives of cloud providers Cloud providers primitives : Using Kubernetes to power our user workloads Kubernetes : Building a purpose-built solution, on top of Virtual Machines or BareMetal servers VMs or BareMetal : Operating our own physical infrastructure and network (aka racking BareMetal machines) Racking Our own users face the same kind of options to deploy and operate their production applications. Most of them will rule out options and as these are irrelevant for most businesses nowadays. 3 4 We believe anybody who is not in the infrastructure business should try to build on top of high-level providers like Koyeb, aka option 1. Our first choice: Kubernetes to the rescue We quickly ruled out racking servers (option 4). We've been in the business of racking and even building BareMetal servers but there are now enough options to pick from, at least for our current needs. In 2021, building software usually involves relying on a cloud service provider. We are no exception to that as we are . before building a cloud service provider on top of other cloud service providers We also ruled out building on top of high-level abstractions of cloud service providers (option 1) since it doesn't provide the control we need. This option would probably allow us to develop features faster, but our platform would suffer from the same product limits as these abstractions. Since our goal is to remove these limits to improve our users' deployment experience, this choice would be counter-productive. With this in mind, we started to build on top of Kubernetes, believing this would be the right abstraction layer to quickly build a reliable platform meeting our requirements. Spoiler alert, after several months we decided to build a purpose-built solution instead... The limits of Kubernetes At Koyeb, we have been operating Kubernetes clusters for two years now and, once we started to support custom user functions and containers, we began to see limits. Here is a broad overview of the limits we started to hit: Let's start with the most common complaint you might hear: it's complex. In our case, we need to be able to extend the core. Even when operating a cloud platform is your business, understanding how Kubernetes works and why it was implemented this way is hard. Complexity: Kubernetes provides some options to isolate multi-tenant serverless workloads. We first worked with , but the performances were disappointing. We decided to explore using Firecracker but running Firecracker on Kubernetes was experimental. Security: gVisor User workloads on Koyeb need to be able to run in multiple zones. Kubernetes doesn't support multi-zone out of the box. Implementing multi-zone with Kubernetes requires deploying a full cluster per zone, with a dedicated control plane for each data center. Global and multi-zone deployments: Kubernetes is known to have limits in terms of the number of nodes in a cluster, which means that multi-cluster support needs to be implemented really quickly when operating at scale. Scalability: Release cycles are really short, which is great but difficult to manage in production. Having to upgrade every two months, which means testing and validating that the upgrades do not break the production quickly becomes a full-time job. Upgradability: Kubernetes clusters tend to have . On each hypervisor, Kubelet uses between 10% and 25% of RAM which is a non-negligible cost. Spoiler alert: We're more around 100MB with our new architecture. Overhead: considerable overhead Kubernetes: a generalist orchestration platform Back in the early 2010s, the hype was more around than Kubernetes as Docker didn't even exist, but Kubernetes is the same kind of project: an all-purpose orchestration platform. OpenStack In our experience as a cloud service provider, when you work with large general-purpose software projects, you tend to hit the same issues: Once you start scaling or trying to develop advanced features on such a platform, you need to fully understand the internals. You spend a lot of energy trying to follow upgrades that might not be relevant to you, break your integrations, and you suffer from decisions you have no control over Don't get me wrong, Kubernetes definitely helped us develop quickly the first generations of the Koyeb engine. The ability to deploy pre-written Helm Charts is a powerful tool to deploy existing software. As a cloud service provider, we knew we would eventually hit the limits and would either need to deep dive into Kubernetes and contribute inside its core or design the Koyeb platform differently... The stack behind Koyeb: Nomad, Firecracker, Kuma on BareMetal... and Kubernetes With all these limits in mind, we started to re-design the platform to support our goals. After some testing, we decided to build our serverless platform using a combination of powerful technologies. Enter Nomad, Firecracker, and Kuma on BareMetal... and Kubernetes. Overall architecture To understand why we combined these technologies, let's take a step back. On the Koyeb platform, there are two key parts: The is composed of the Go APIs behind and related databases. It receives the API calls and handles all the logic to effectively deploy the app on the Koyeb Data Plane. It needs to run 24x7x365 with requirements pretty similar to those of users' on Koyeb. Koyeb Control Plane app.koyeb.com The hosts the deployed users' applications. It's the core of the platform, it needs to support running millions of containers across thousands of hypervisors. It's the most demanding part of the platform. Koyeb Data Plane On each hypervisor of the data plane, there are 4 major elements: The , Firecracker, effectively executing the applications, ensuring multi-tenants workloads are isolated and secured in a performant fashion. container runtime and virtualization technology The , composed of Kuma and Envoy, provides the service mesh and discovery engine which enable seamless private networking between services. networking technology The , Nomad, which performs all provisioning operations when a deployment happens or scaling events happen. orchestration engine The , CoreOS on High-End BareMetal servers, where the applications run with the lowest possible overhead. operating system and servers And the Control Plane runs on... Kubernetes. As we cannot run the Koyeb Control Plane on Koyeb to avoid circular dependencies, we run the APIs on top of Kubernetes. Diving into Nomad, Firecracker, and Kuma Nomad for Container Orchestration is a workload orchestration engine that can be used to manage containers and non-containerized applications across clouds. Nomad lets us deploy and manage the containerized applications that are running on the platform. It provisions Firecracker microVMs and communicates with containers to manage and run the containers. Nomad We selected Nomad to do the orchestration because compared to Kubernetes it is: Easily extendable. Runtimes can easily be customized and integrating with Firecracker was seamless using a custom nomad driver. Simpler to use. The scope of the project is way smaller, it doesn't cover everything from runtime to networking and that's just what we needed. Multi-zone out of the box, simplifying how we make serverless workloads global. Easy to implement control plane at the region level, reducing the overhead of per-region Kubernetes clusters. Has a smaller footprint and is easily scalable. Firecracker for Virtualization is a lightweight virtualization technology that was purpose-built to securely and efficiently run multi-tenant serverless workloads. We needed a virtualization technology that would keep our clients' workloads secure and isolated and provide fast boot-up times for new instances without compromising performances. That's exactly what Firecracker does. Firecracker We've written about in a previous blog post if you want to dive deeper. Firecracker MicroVMs The key benefits are: Multi-tenancy security and isolation, High density on bare metal machines thanks to low overhead, Fast boot-times for autoscaling microservices and functions There are a few more benefits that we go over in detail in our post . 10 Reasons Why We Love Firecracker MicroVMs Kuma for Service Mesh Control Plane is an open-source service mesh built on top . It is a control plane that connects distributed services running on Kubernernets and VMs. Kuma Envoy We built Koyeb to host distributed and microservice architectures. Distributed architectures need a way to connect the services that run within them. A service mesh is the dedicated infrastructure layer that is responsible for the secure, quick, and reliable communication between services in containerized and microservice applications. For more information about service meshes, check out our post on . service meshes and microservices Kuma is a rare service mesh that is not tightly coupled to its workload scheduler. This distinguishes it from alternative service meshes like Linkerd and Istio, which only work on Kubernetes. Moreover, Kuma supports multi-zone and cloud-agnostic out of the box. The Kuma control plane allows us to secure, observe, and manage the microservice network. Next to each service instance on Koyeb is an Envoy Proxy. These proxies form the data plane, which connects to the Kuma control plane. Proxies are configured from the control plane, meaning the hostnames your instances should know for DNS, their retry policies, and more are all communicated to the proxy from the control plane. Kuma and Envoy provide us with a great option for private networking, especially because they: Are platform-agnostic, Support multi-region and multi-cloud deployments, Can run in a zero-trust network environment thanks to native TLS encryption and mTLS authentication, Provide natively secure communication between services. The Koyeb Serverless Platform is Purpose-Built We built Koyeb to be a secure and powerful serverless platform that is developer-friendly. Providing this in a secure and resource-efficient manner. Moving from Kubernetes to a customized stack of Nomad, Firecracker, and Kuma enables us to provide the serverless experience we envision for the serverless future. serverless experience entails running multi-tenant workloads on bare metal servers You can use Koyeb to host web apps and services, Docker containers, APIs, event-driven functions, cron jobs, and more. At the moment, the platform has built-in Docker container deployment and git-driven deployment is on the way! Discover the serverless experience and by today. Also, feel free to join us over on . signing up Slack And if you want to help build a serverless cloud service provider and dive deeper into these technologies, ! we’re hiring

Building The Koyeb Serverless Engine: Why We Moved From Kubernetes to Nomad, Firecracker, and Kuma

, product leader of a family of stablecoins, Samecoin, spoke to me recently in order to share his expert knowledge by answering some of the questions about the stablecoin sector in the crypto niche. Kurt Farrugia 1) What makes stablecoins different from regular cryptocurrencies? Stablecoins are supposed to hold their value---rather than “moon” like other coins. Crypto has been a huge investment vehicle, and it’s helped transform my finances---but BTC, ETH, and other altcoins simply aren’t stable currencies. They’re investments. With stablecoins, you can actually start spending digital currencies on all sorts of things without having to worry your transaction’s value will be worth a completely different amount tomorrow. 2) Do you consider stablecoins a better version of crypto and which is the most important? Stablecoins aren’t “better”---but they are different. There’s a place for BTC and all those other cryptocurrencies, but in order for people to start using crypto for its original intended use, and be used every day for normal purchases, stablecoins are an important step towards that future. 3) How do you respond to claims that stablecoins are not as good as BTC due to the element of centralization? It depends on what you’re looking for. So far, stablecoins have been centralized as they are usually backed by a reserve made up of bank accounts or assets held by companies or individuals. However, new decentralized stablecoins are emerging, some of which are fully decentralized, as all operations and reserves can be verified by anyone by simply checking the smart contract and the value locked in it. 4) Are all stablecoins backed by fiat, or are there fiat-independent ones? There are loads of different types of stablecoins. Some are backed by fiat, while others are backed by other cryptocurrencies or even commodities like gold. In some cases, they are backed by other stablecoins in the market, with algorithms and operations that help distribute the risk and volatility. 5) Which are better in your opinion? Instead of being “backed” by an individual fiat or individual asset, the best stablecoins are pegged to a basket of assets to help ensure absolute stability. These can be tied to a fiat coin value, without being backed by that fiat’s own reserves (yes, it’s doable with the emerging technologies) and are more resilient to fluctuations. Most importantly, it is fully transparent and can be fully decentralized. With DeFi integrations, what can get better? 6) Do you think that stablecoins might push cryptos like Bitcoin and Ethereum out of business, eventually? Probably not. But there’s absolutely huge space for growth in the stablecoin world. BTC and ETH are probably here to stay, but they’re mostly seen as investments rather than currency you can actually spend on everyday needs. 7) Stablecoins are facing some problems of their own. Can you tell us about the biggest issues? There have been a few regulatory issues with some of the biggest stablecoins. There’s also a lack of transparency in some of the industry, and questions still remain when it comes to some stablecoins like USDT. The objective should be to bring a stablecoin system that is focused on adoption, integrating with various websites, applications, services, and eventually also your local supermarkets or restaurants, bringing real-world utility. 8) Do you think that these problems are preventing them from becoming more adopted? Absolutely, people need a stablecoin they can trust. Unfortunately, crypto has not been as widely adopted as we expected in our everyday lives. The vast majority of people still do not know how to buy crypto and where and how to use it. We need it simplified. 9) Let’s talk about Samecoin. How is the project contributing to the niche? is an entire ecosystem, made up of cryptocurrencies and the products which will drive up their adoption. This is the advantage of Samecoin; it’s how we focus on the user experience, translating and transforming the complexities of crypto for the billions of people who know about crypto but are too afraid to get to know it and use it. Samecoin We hope this will make Samecoin one of the first to achieve adoption in that untapped market. 10) How can Samecoin solve the problems that stablecoins are struggling with? Samecoin is a completely transparent ecosystem that anyone can investigate, unlike most others in the stablecoin world. It also helps bring DeFi and stablecoins to the masses with a spendable currency that’s easy to understand and easy to use. That’s why I’m passionate about Samecoin. We are building an ecosystem of cryptocurrencies and products that simplify crypto for the everyday person, without taking the shortcuts and risks of centralization. SameUSD, and SameEuro are the stable and decentralized cryptocurrencies that people and businesses have been waiting for. The asset reserves are fully transparent and visible in the smart contracts. They have been audited, verified, and passed by Certik. SamePay and SameID on the other hand, will be the payments and accounts products that will help drive up adoption, by giving people and businesses easy-to-use products that bring crypto into their lives. Support for businesses and merchants helps create a growing ecosystem of payment solutions that can make crypto’s vision finally real. Upon launch on the 8th of June, 2021, Samecoin beta app amassed $2 million TVL within the first hour. Disclosure: The author does not have any vested interest in the projects mentioned above. Disclaimer: The opinions in this article belong to the author alone. Nothing in this article constitutes investment advice. Please conduct your own thorough research before making any investment decisions.

Diving into Stablecoins: Interview with Kurt Farrugia

BinaryX announced today the highly anticipated launch of their first Initial Game Offering(IGO) of the year, Project Matthew. SINGAPORE, Singapore, February 29th, 2024, Chainwire/ Project Matthew is a space-building simulation game where players get to experience what it is like to work in space, officially known as the planet Matthew in the game. In the game, players work not as astronauts, but as builders and engineers that make space livable for a new civilization. Own ERC-404 NFT Tokens And Exchange For In-game Land Players have the opportunity to own fractionalized that can be converted to ownership of in-game land, weapons, and other materials. ERC-404 tokens provide users with ample liquidity while also opening up different possibilities for their future ownership, including shared ownership of land, buildings and resources. ERC-404 NFTs The game, with its innovative use of ERC-404 tokens, is expected to promote collaboration, a sense of community, and open the game to a wider audience without compromising the exclusivity of in-game NFT ownership. The Story of Project Matthew: A Space Building Simulator Humans discover a not-so-distant planet called Matthew with the potential to house living beings. As pioneers of space living, players will have the opportunity to design and manage their very own industrial plants, mine for resources, build robots that can help to work in the plants as well as fight other intergalactical forces, and construct their very own space city from the ground up. Lead A Virtual Army Players can recruit a robot army with different skills and abilities, and challenge enemies on the battlefield to earn massive rewards. The battlefield is divided into different levels of difficulties. The more difficult the level, the greater the rewards. Explore Space Exploration is one of the primary ways to obtain rewards and resources in Project Matthew. The gameplay features a collection of neighboring and faraway planets waiting to be explored. “Project Matthew is our first IGO project for the first half of 2024. We wanted our players to have the opportunity to experience living and working in space, and this game is the first ever simulation game in the Web3 space to offer a semi-realistic experience for that”, said Julio, Head of Investments at BinaryX. “We’re excited and geared up to support more quality games like Project Matthew on our IGO platform in 2024.” Participate in the IGO here About BinaryX BinaryX is a leading Web3 gaming platform dedicated to creating immersive and engaging gaming experiences like and . Built on the BSC Chain, BinaryX leverages the latest technologies in blockchain and AI to provide over 100k players with unique opportunities to own in-game assets, participate in decision-making processes, and experience the true potential of Web3 gaming. Ai Hero Pancake Mayor For more details and information about BinaryX, please visit www.binaryx.pro Find us on social media: BinaryX BinaryX Medium Contact Kora marketing@binaryx.pro This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here.

KURA TECHNOLOGIES

kura.tech

-- employees

Since n.d.

COMPANY RANKING

KURA TECHNOLOGIES

EVERGREEN INDEX #390

Representing this company? Click here to claim and customize this page!

Kura Technologies WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

KURA TECHNOLOGIES

kura.tech

-- employees

Since n.d.

COMPANY RANKING

KURA TECHNOLOGIES

EVERGREEN INDEX #390

RELATED COMPANIESRANK

STORIES ABOUT KURA TECHNOLOGIES

Representing this company? Click here to claim and customize this page!

Kura Technologies WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!