Kalto

kalto

Using Scanned Mesh Data for Auto-Digitized 3D Modeling: Methodology

Here are the simplest steps needed to create a deployment from your local GIT repository to a remote server. Goal When you run , deploy the latest branch distribution to the server without preparations on the server-side. git push master Introduction In addition to writing code, indie developers need to think about the continuous delivery of their services. And often setting up a deployment pipeline is a non-trivial task: you need to set up Jenkins, Kubernetes, Git Hooks, and many other technologies. In addition to the fact that they require such a valuable programmer resource as the time needed for learning them, they also consume hardware resources. But what if I have a Linux server for several gigabytes where I want to place, for example, my small blog with zero configs? gradle-ssh-plugin As the plugin author, I promise that you will not find an easier way to deploy than this. Prerequisites If you have a remote Linux server you must have the file. id_rsa When setting up a remote Linux server, you’ll need to decide upon a method for securely connecting to it. While passwords are one way of verifying a user’s identity, passwords have multiple vulnerabilities and can be cracked by a brute force attack. ecure ell keys — better known as — are often used instead of passwords, as they offer a more secure method of connecting to remote Linux servers. As part of the Secure Shell cryptographic network protocol, SSH keys also enable users to securely perform network services over an unsecured network, such as delivering text-based commands to a remote server or configuring its services. S Sh SSH keys If you don’t have this key let’s create it - run this commands on your local machine: ssh-keygen -m PEM -t rsa -b 2048 chmod 400 ~/.ssh/id_rsa ssh-copy-id -i .ssh/id_rsa.pub root@host.domain Now you need to copy the file ~/.ssh/ from the local machine into the root of your project which you need to deploy. id_rsa 🎯 Quick start In root project file: build.gradle.kts plugins {
    id("online.colaba.ssh") version "1.8.17"
}
group = "online.colaba" That's all! After running a specific grade task your code distribution will be delivered to the remote server. This task will copy folders & files from local machine to remote host folder ~/${project.name}/... You can set host, or it will computed from (example above) project.group tasks {
    scp { 
        host = "colaba.online"
    }
} Run deploy task: gradle scp 🔮 Customization: Register new task in your : build.gradle.kts register("customSshTask", Ssh::class) {
   host = "my-domain.com"
   user = "root"
   gradle = true
   frontend = false
   docker = true
   nginx = true
   directory = "distribution"
   run = "cd ${project.name} && echo \$PWD"
} Run this task: gradle customSshTask 🌀 Available Gradle tasks from plugin: ssh By default you have preconfigured tasks: - all options are by default ( ) ssh disabled false - all options are by default ( ) scp enabled true - copy needed files to remote server in every subproject ssh-gradle gradle - copy files to remote server ssh-docker docker - copy file to remote server in every subproject ssh-jars ${subproject}/build/libs/___.jar Example of tasks that will become available for your project: There will be as many tasks as Gradle subprojects. - copy distribution -file to remote server ssh-backend backend *.jar - copy folder to remote server ssh-frontend frontend - copy folder to remote server ssh-nginx nginx ... Name of service for all tasks equals to ${project.name} 📋 Project's structure example There could be as many backends as you need. project
   |-[backend]
              | - [src/main/java/build/libs]/*.jar
              | - Dockerfile
              | - Dockerfile.dev
              | - docker-compose.yml
              | - docker-compose.dev.yml
              | - ...
   |-[backend-2]
              | - [src/main/koltin/build/libs]/*.jar
              | - ...
   |-[backend-3]
              | - [src/main/scala/build/libs]/*.jar
              | - ...
   |-[frontend]
              | - docker-compose.yml
              | - ...
   |-[nginx]
              | - ...
   |-[postgres]
              | - [backups]
              | - ...
   |-[elastic]
              | - ...
   |-[static]
              | - ...
   |-[gradle]
              | - [wrapper]
   |- gradlew
   |- gradlew.bat
   |- docker-compose.yml
   |- ... A more detailed description of this plugin and examples will be in my next articles soon.

The Easiest Way To Deploy With Gradle

Hey Hackers! Have you seen the new yellow section on the homepage called ✨ Boosted Stories ✨? Let us introduce you to HackerNoon’s newest feature—just in time for gift-giving season! What’s a ✨ Story Boost ✨? As the name suggests, Story Boost is a feature designed to amplify your content by showcasing it on: The HackerNoon Homepage: Position your story as valuable and noteworthy while boosting your reputation as a thought leader.
The HackerNoon Newsletter: Gain exposure to an engaged audience of 350,000+ tech and business enthusiasts. These prime placements not only increase your visibility but also help establish your personal brand, showcase professional expertise, and build lasting credibility. Not bad, right? Meet The HackerNoon Newsletter (yes, the main one!) Your daily dose of Must-Read Stories curated by HackerNoon editors, delivered straight to your inbox daily, at noon Mountain Time. This newsletter also includes:✨Poll of the Week:community insights rule!✨On this Day:historical events, tech news & tech stories throwback✨Feature Job Postings:anybody looking for a job?✨The Market At a Glance: tickers & prices from our trending companies & coins database ✨ Translations: The newsletter is now available in over 70 languages! Check out all the editions here or subscribe today. How to Boost Your Story It’s easy! Follow these steps to give your story the visibility it deserves: 1. Open any of your stories and click "Boost this story" atop the title to unlock the feature. 2. After reading HackerNoon's story boost process, click "Boost this story" to pay. 3. Enter your card info and hit “Pay Now”. Different payment options may be available in different countries. 4. After payment, you'll be directed to the receipt page. Make sure the correct story is boosted, if not, please contact HackerNoon support! 5. On the same story page, you can click the same button to track the story boost progress. 6. You'll be updated as the story is being boosted on both the homepage and HackerNoon Newsletter. Alternatively, visit the cart system to add a Story Boost to your cart. Once you’ve paid, select the story you want to boost, and voilà! What Our Users Have to Say About Story Boosts Carin Lagerstedt, CMO Remotive Labs Hackernoon’s Boosted Story feature helps us gain visibility among tech enthusiasts and automotive developers. It’s a great way to spotlight how our tools empower software development and reduce dependencies in the automotive industry. Sal Kimmich, Technical Director, GadflyAI I’m thrilled to receive a boost on HackerNoon! It helps me to raise stories that have ongoing impact, like the discussion around transparency and authenticity in open-source AI. HackerNoon’s platform has been instrumental in amplifying these important conversations, but boosts make it possible to amplify the most urgent stories on cybersecurity regulation, directly to technology leaders, in real time. Noah Kalson, Director of Brand Marketing, Bright Data The Boosted Story feature is the best way to highlight our brand's articles on HackerNoon and reach new and engaged readers. HackerNoon has always been a platform I trust for its quality and innovation, and this feature is no exception. They truly understand the needs of content creators and make it so easy to share our content with the right audience. Examples of Boosted Stories 1. Exploring the RIF Token: Bringing Utility to Bitcoin https://hackernoon.com/exploring-the-rif-token-bringing-utility-to-bitcoin?embedable=true 2. Building a Data Analytics Platform to Streamline the Temporary Labor Sector https://hackernoon.com/building-a-data-analytics-platform-to-streamline-the-temporary-labor-sector?embedable=true 3. Bypass GPTZero AI Detection Every Time with PassMe.AI – The Ultimate AI Humanizer https://hackernoon.com/bypass-gptzero-ai-detection-every-time-with-passmeai-the-ultimate-ai-humanizer?embedable=true Learn more about Story Boost here! Hey Hackers! Have you seen the new yellow section on the homepage called ✨ Boosted Stories ✨? Boosted Stories Let us introduce you to HackerNoon’s newest feature—just in time for gift-giving season! What’s a ✨ Story Boost ✨? What’s a ✨ Story Boost ✨? As the name suggests, Story Boost is a feature designed to amplify your content by showcasing it on: The HackerNoon Homepage: Position your story as valuable and noteworthy while boosting your reputation as a thought leader. The HackerNoon Newsletter: Gain exposure to an engaged audience of 350,000+ tech and business enthusiasts. The HackerNoon Homepage : Position your story as valuable and noteworthy while boosting your reputation as a thought leader. The HackerNoon Homepage The HackerNoon Homepage The HackerNoon Newsletter : Gain exposure to an engaged audience of 350,000+ tech and business enthusiasts . The HackerNoon Newsletter The HackerNoon Newsletter engaged audience of 350,000+ tech and business enthusiasts These prime placements not only increase your visibility but also help establish your personal brand, showcase professional expertise, and build lasting credibility. Not bad, right? Meet The HackerNoon Newsletter (yes, the main one!) Meet The HackerNoon Newsletter (yes, the main one!) HackerNoon Newsletter (yes, the main one!) Your daily dose of Must-Read Stories curated by HackerNoon editors, delivered straight to your inbox daily, at noon Mountain Time. This newsletter also includes:✨Poll of the Week:community insights rule!✨On this Day:historical events, tech news & tech stories throwback✨Feature Job Postings:anybody looking for a job?✨The Market At a Glance: tickers & prices from our trending companies & coins database Must-Read Stories curated by HackerNoon editors, delivered straight to your inbox daily, at noon Mountain Time. This newsletter also includes:✨Poll of the Week:community insights rule!✨On this Day:historical events, tech news & tech stories throwback✨Feature Job Postings:anybody looking for a job?✨ The Market At a Glance: Must-Read Stories curated by HackerNoon editors, delivered straight to your inbox daily, at noon Mountain Time. This newsletter also includes: ✨Poll of the Week: community insights rule! ✨On this Day: historical events, tech news & tech stories throwback ✨Feature Job Postings: anybody looking for a job? ✨ ✨ Poll of the Week : Poll of the Week ✨ On this Day : On this Day ✨ Feature Job Postings : Feature Job Postings trending companies coins ✨ Translations : The newsletter is now available in over 70 languages! Translations Translations Check out all the editions here or subscribe today . all the editions here subscribe today How to Boost Your Story It’s easy! Follow these steps to give your story the visibility it deserves: 1. Open any of your stories and click "Boost this story" atop the title to unlock the feature. 1. Open any of your stories and click "Boost this story" atop the title to unlock the feature. 2. After reading HackerNoon's story boost process, click "Boost this story" to pay. 2. After reading HackerNoon's story boost process, click "Boost this story" to pay. 3. Enter your card info and hit “Pay Now”. Different payment options may be available in different countries. 3. Enter your card info and hit “Pay Now”. Different payment options may be available in different countries. 4. After payment, you'll be directed to the receipt page. Make sure the correct story is boosted, if not, please contact HackerNoon support! 4. After payment, you'll be directed to the receipt page. Make sure the correct story is boosted, if not, please contact HackerNoon support ! contact HackerNoon support 5. On the same story page, you can click the same button to track the story boost progress. 5. On the same story page, you can click the same button to track the story boost progress. 6. You'll be updated as the story is being boosted on both the homepage and HackerNoon Newsletter. 6. You'll be updated as the story is being boosted on both the homepage and HackerNoon Newsletter. Alternatively, visit the cart system to add a Story Boost to your cart. Once you’ve paid, select the story you want to boost, and voilà! Alternatively, visit the cart system to add a Story Boost to your cart. Once you’ve paid, select the story you want to boost, and voilà! cart system What Our Users Have to Say About Story Boosts What Our Users Have to Say About Story Boosts Carin Lagerstedt, CMO Remotive Labs Hackernoon’s Boosted Story feature helps us gain visibility among tech enthusiasts and automotive developers. It’s a great way to spotlight how our tools empower software development and reduce dependencies in the automotive industry. Hackernoon’s Boosted Story feature helps us gain visibility among tech enthusiasts and automotive developers. It’s a great way to spotlight how our tools empower software development and reduce dependencies in the automotive industry. Sal Kimmich, Technical Director, GadflyAI I’m thrilled to receive a boost on HackerNoon! It helps me to raise stories that have ongoing impact, like the discussion around transparency and authenticity in open-source AI. HackerNoon’s platform has been instrumental in amplifying these important conversations, but boosts make it possible to amplify the most urgent stories on cybersecurity regulation, directly to technology leaders, in real time. I’m thrilled to receive a boost on HackerNoon! It helps me to raise stories that have ongoing impact, like the discussion around transparency and authenticity in open-source AI. HackerNoon’s platform has been instrumental in amplifying these important conversations, but boosts make it possible to amplify the most urgent stories on cybersecurity regulation, directly to technology leaders, in real time. Noah Kalson, Director of Brand Marketing, Bright Data The Boosted Story feature is the best way to highlight our brand's articles on HackerNoon and reach new and engaged readers. HackerNoon has always been a platform I trust for its quality and innovation, and this feature is no exception. They truly understand the needs of content creators and make it so easy to share our content with the right audience. The Boosted Story feature is the best way to highlight our brand's articles on HackerNoon and reach new and engaged readers. HackerNoon has always been a platform I trust for its quality and innovation, and this feature is no exception. They truly understand the needs of content creators and make it so easy to share our content with the right audience. Examples of Boosted Stories 1. Exploring the RIF Token: Bringing Utility to Bitcoin 1. Exploring the RIF Token: Bringing Utility to Bitcoin https://hackernoon.com/exploring-the-rif-token-bringing-utility-to-bitcoin?embedable=true https://hackernoon.com/exploring-the-rif-token-bringing-utility-to-bitcoin?embedable=true 2. Building a Data Analytics Platform to Streamline the Temporary Labor Sector 2. Building a Data Analytics Platform to Streamline the Temporary Labor Sector https://hackernoon.com/building-a-data-analytics-platform-to-streamline-the-temporary-labor-sector?embedable=true https://hackernoon.com/building-a-data-analytics-platform-to-streamline-the-temporary-labor-sector?embedable=true 3. Bypass GPTZero AI Detection Every Time with PassMe.AI – The Ultimate AI Humanizer 3. Bypass GPTZero AI Detection Every Time with PassMe.AI – The Ultimate AI Humanizer https://hackernoon.com/bypass-gptzero-ai-detection-every-time-with-passmeai-the-ultimate-ai-humanizer?embedable=true https://hackernoon.com/bypass-gptzero-ai-detection-every-time-with-passmeai-the-ultimate-ai-humanizer?embedable=true Learn more about Story Boost here! Learn more about Story Boost here ! here

Reach More Readers! BOOST Your Story on HackerNoon

ብዙ አንባቢዎችን ይድረሱ! በ HackerNoon ላይ ታሪክዎን ያሳድጉ

Erreichen Sie mehr Leser! Boosten Sie Ihre Geschichte auf HackerNoon

¡Llega a más lectores! POTENCIA tu historia en HackerNoon

Tavoita lisää lukijoita! TEHOSTA tarinaasi HackerNoonissa

Հասնել ավելի շատ ընթերցողների: ԱՇԽԱՏԵԼ Ձեր պատմությունը HackerNoon-ում

Raggiungi più lettori! POTENZIA la tua storia su HackerNoon

より多くの読者にリーチしましょう！HackerNoonであなたのストーリーを宣伝しましょう

Tongava Mpamaky bebe kokoa! TOROHEVITRA ny tantaranao amin'ny HackerNoon

Fihlelela Babadi ba Bantši! BOOST Pale ya Gago go HackerNoon

نور لوستونکو ته ورسیږئ! په هیکرنون کې خپله کیسه ښه کړئ

Shikira Abasomyi Benshi! SHAKA inkuru yawe kuri HackerNoon

Ko'proq o'quvchilarga murojaat qiling! HackerNoon-da hikoyangizni BOOST

TEL AVIV, ISRAEL, December 3rd, 2024/CyberNewsWire/--With Sweet, customers can now unify detection and response for applications, workloads, and cloud infrastructure. Sweet Security today announced the release of its unified Cloud Native Detection and Response platform, designed to transform the way organizations protect their cloud environments in real time. Sweet’s platform integrates the capabilities of Application Detection and Response (ADR), Cloud Detection and Response (CDR), and Cloud Workload Protection Platform (CWPP) into one comprehensive solution. This innovative approach delivers unparalleled detection and response capabilities, unifying insights from every layer of the cloud stack. “The Sweet team has worked tirelessly to build a platform that not only meets the needs of today’s security landscape but also anticipates the challenges of tomorrow,” said Dror Kashti, CEO of Sweet. “With our supercharged detection and response capabilities, we’re providing organizations with the tools they need to secure their cloud environments faster and more effectively than ever before.” In an era where security teams face an average dwell time of 10 days to resolve cloud incidents due to lack of context, Sweet Security empowers organizations to put detection and response at the forefront of their cloud defense strategy, providing the ability to act on threats as they happen and ensuring an incident never becomes a breach. When detection and response is siloed between application, workload, and cloud infrastructure levels, it results in fragmented alerts that require security teams to piece together in order to understand the scope of the attack. "Sweet Security's Cloud Native Detection and Response platform has been a game changer for us,” said Shai Sivan, CISO at Kaltura. “The ability to unify visibility across both application activity and infrastructure changes has been crucial in gathering the forensic data we need. With average detection times of just 30 seconds and the ability to respond to incidents in 2-5 minutes, Sweet has decreased our Mean Time to Resolution (MTTR) by an impressive 90%, allowing us to respond faster and more effectively." In addition to unified detection and response, Sweet Security’s platform is equipped with a suite of next-generation capabilities that are powered by GenAI, including: Unified Cloud Visibility
Vulnerability Management
Runtime CSPM
Identity Threat Detection & Response (ITDR) James Berthoty, cloud expert and owner of Latio Tech, recently commented on Sweet Security’s unified solution, stating, “Cloud Application Detection & Response is all about combining cloud, workload, and application layer contexts into a single storyline to make incident response in the cloud finally a reality. For too long we've been content with mediocre visibility into our most important assets. Sweet is on the frontlines of changing that by giving teams the cross-layer visibility they've needed for years.” Book a meeting with Sweet Security at AWS re:Invent 2024 in Las Vegas here. About Sweet Security Specializing in Cloud Native Detection & Response (D&R), Sweet Security protects cloud environments in real time. Founded by the IDF's former CISO, Sweet's solution unifies threats insights from cloud applications, workloads and infrastructure. Leveraging a lean, eBPF-based sensor and deep behavioral analysis, Sweet analyzes anomalies, generating vital insights on incidents, vulnerabilities, and non-human identities. Its GenAI-infused technology cuts through the noise and delivers actionable recommendations on critical, real-time cloud risks. Privately funded, Sweet is backed by Evolution Equity Partners, Munich Re Ventures, Glilot Capital Partners, CyberArk Ventures and an elite group of angel investors. For more information, please visit http://sweet.security. Contact Elizabeth Safran Looking Glass Public Relations for Sweet Security 381031@email4pr.com This story was distributed as a release by Cybernewswire under HackerNoon’s Business Blogging Program. Learn more about the program here TEL AVIV, ISRAEL, December 3rd, 2024/CyberNewsWire/--With Sweet, customers can now unify detection and response for applications, workloads, and cloud infrastructure. Sweet Security today announced the release of its unified Cloud Native Detection and Response platform, designed to transform the way organizations protect their cloud environments in real time. Sweet Security Sweet Security Sweet’s platform integrates the capabilities of Application Detection and Response (ADR), Cloud Detection and Response (CDR), and Cloud Workload Protection Platform (CWPP) into one comprehensive solution. This innovative approach delivers unparalleled detection and response capabilities, unifying insights from every layer of the cloud stack. “The Sweet team has worked tirelessly to build a platform that not only meets the needs of today’s security landscape but also anticipates the challenges of tomorrow,” said Dror Kashti, CEO of Sweet. “The Sweet team has worked tirelessly to build a platform that not only meets the needs of today’s security landscape but also anticipates the challenges of tomorrow,” said Dror Kashti, CEO of Sweet. “With our supercharged detection and response capabilities, we’re providing organizations with the tools they need to secure their cloud environments faster and more effectively than ever before.” “With our supercharged detection and response capabilities, we’re providing organizations with the tools they need to secure their cloud environments faster and more effectively than ever before.” In an era where security teams face an average dwell time of 10 days to resolve cloud incidents due to lack of context, Sweet Security empowers organizations to put detection and response at the forefront of their cloud defense strategy, providing the ability to act on threats as they happen and ensuring an incident never becomes a breach. When detection and response is siloed between application, workload, and cloud infrastructure levels, it results in fragmented alerts that require security teams to piece together in order to understand the scope of the attack. "Sweet Security's Cloud Native Detection and Response platform has been a game changer for us,” said Shai Sivan, CISO at Kaltura. "Sweet Security's Cloud Native Detection and Response platform has been a game changer for us,” said Shai Sivan, CISO at Kaltura. “The ability to unify visibility across both application activity and infrastructure changes has been crucial in gathering the forensic data we need. With average detection times of just 30 seconds and the ability to respond to incidents in 2-5 minutes, Sweet has decreased our Mean Time to Resolution (MTTR) by an impressive 90%, allowing us to respond faster and more effectively." “The ability to unify visibility across both application activity and infrastructure changes has been crucial in gathering the forensic data we need. With average detection times of just 30 seconds and the ability to respond to incidents in 2-5 minutes, Sweet has decreased our Mean Time to Resolution (MTTR) by an impressive 90%, allowing us to respond faster and more effectively." In addition to unified detection and response, Sweet Security’s platform is equipped with a suite of next-generation capabilities that are powered by GenAI, including: Unified Cloud Visibility Vulnerability Management Runtime CSPM Identity Threat Detection & Response (ITDR) Unified Cloud Visibility Vulnerability Management Runtime CSPM Identity Threat Detection & Response (ITDR) James Berthoty, cloud expert and owner of Latio Tech, recently commented on Sweet Security’s unified solution, stating, “Cloud Application Detection & Response is all about combining cloud, workload, and application layer contexts into a single storyline to make incident response in the cloud finally a reality. For too long we've been content with mediocre visibility into our most important assets. Sweet is on the frontlines of changing that by giving teams the cross-layer visibility they've needed for years.” “Cloud Application Detection & Response is all about combining cloud, workload, and application layer contexts into a single storyline to make incident response in the cloud finally a reality. For too long we've been content with mediocre visibility into our most important assets. Sweet is on the frontlines of changing that by giving teams the cross-layer visibility they've needed for years.” Book a meeting with Sweet Security at AWS re:Invent 2024 in Las Vegas here . here here About Sweet Security Specializing in Cloud Native Detection & Response (D&R), Sweet Security protects cloud environments in real time. Founded by the IDF's former CISO, Sweet's solution unifies threats insights from cloud applications, workloads and infrastructure. Leveraging a lean, eBPF-based sensor and deep behavioral analysis, Sweet analyzes anomalies, generating vital insights on incidents, vulnerabilities, and non-human identities. Its GenAI-infused technology cuts through the noise and delivers actionable recommendations on critical, real-time cloud risks. Privately funded, Sweet is backed by Evolution Equity Partners, Munich Re Ventures, Glilot Capital Partners, CyberArk Ventures and an elite group of angel investors. For more information, please visit http://sweet.security . http://sweet.security http://sweet.security Contact Elizabeth Safran Looking Glass Public Relations for Sweet Security 381031@email4pr.com This story was distributed as a release by Cybernewswire under HackerNoon’s Business Blogging Program. Learn more about the program here This story was distributed as a release by Cybernewswire under HackerNoon’s Business Blogging Program. Learn more about the program here here here

Sweet Security Unveils First Unified Detection And Response Platform

Sweet Security onthul eerste verenigde opsporing- en reaksieplatform

Sweet Security presenta la primera plataforma unificada de detección y respuesta

Sweet Security julkistaa ensimmäisen yhtenäisen tunnistus- ja vastausalustan

Sweet Security presenta la prima piattaforma unificata di rilevamento e risposta

スウィートセキュリティ、初の統合検出・対応プラットフォームを発表

Sweet Security биринчи бирдиктүү аныктоо жана жооп берүү платформасын ачат

Sweet Security ја претстави првата обединета платформа за откривање и одговор

मीठो सुरक्षा पहिलो एकीकृत पत्ता लगाउने र प्रतिक्रिया प्लेटफर्म अनावरण

Sweet Security представляет первую унифицированную платформу обнаружения и реагирования

Umutekano uryoshye ugaragaza uburyo bwa mbere bwo guhuriza hamwe no gusubiza

Inotapira Chengetedzo Inovheneka Yekutanga Yakabatana Kuonekwa uye Mhinduro Yepuratifomu

ஸ்வீட் செக்யூரிட்டி முதல் ஒருங்கிணைந்த கண்டறிதல் மற்றும் பதில் தளத்தை வெளியிடுகிறது

DreamLLM: What We Can Conclude From This Comprehensive Framework?

Advancing the Detection of Ultralight Vector Dark Matter through Novel Analytical Strategies

Hey Hackers! Cropler has been nominated in HackerNoon's Startups of The Year 2024  awards in Warsaw, Poland under the Emerging Tech category and others. Please vote for us here: 1️⃣ https://hackernoon.com/startups/europe/europe-warsaw-poland 2️⃣ https://hackernoon.com/startups/industry/iot 3️⃣ https://hackernoon.com/startups/industry/hardware 4️⃣ https://hackernoon.com/startups/industry/climatetech Meet Cropler Cropler is the world’s first real-time remote photo plant monitoring technology that is about to become an AI-powered digital agronomical assistant. Cropler uses smart AI-powered mobile agricameras to help farmers and agribusinesses identify the key phenophases, diseases, and inputs efficiently to perform necessary agricultural operations in a timely manner. The Winning Team 1️⃣ Nik Bedunkevich
LinkedIn
Founder & CEO
With over 10 years in software and 7+ years in hardware development, Nik's leadership in precision agriculture, backed by his experience at OneSoil, where he co-founded and led hardware development, has been instrumental. His background in Radio Engineering and his ability to integrate data exchange and soil analysis sensors have enabled Cropler to push the boundaries of AI-driven agricultural solutions.2️⃣ Marie SinkovaLinkedIn
CMO
Marie’s 6+ years in marketing, including her impactful work at MEDvidi and EduDo, have significantly boosted our brand’s visibility. Her ability to develop engaging content and manage successful teams has increased revenue and brand recognition while minimizing negative reviews. Her strategic approach in various marketing areas has laid a solid foundation for Cropler development.3️⃣ Andrii ZbrytskyiLinkedIn
BDM
With over 17 years in the agritech area, Andri Zbrytskyi has witnessed firsthand the ins and outs of clients resulting in better client/corporate relations, sales and even more business development for the company. AI agriculture solutions have found strategic partnerships thanks to his efforts and these have increased expanding the market base.4️⃣ Xenia NiaborskayaLinkedIn
Head of Design
Xenia’s experience of more than 4 years in UX/UI design has played a key role in ensuring that our user interfaces work as expected and look good as well. Her capability in business-centered design and the process of user-centered design has improved the usability of our platform tremendously.5️⃣ Eugene VoronetsLinkedIn
Tech Adviser
Eugene’s 10+ years in backend development, particularly his expertise in high-load systems, ensures the robustness and scalability of our technology. His work at OneSoil, building APIs and fostering collaborations with industry leaders like John Deere and AWS, has played a critical role in shaping Cropler’s technical infrastructure. What Cropler does differently We aim to revolutionize agriculture with Cropler's global plant monitoring technology, which uses AI-powered agricameras and remote sensing data to provide farmers, agronomists, and agribusinesses with predictive field insights and global yield analytics for optimized crop production. Cropler’s unique, AI-powered plant health data allows farmers and agribusinesses to address issues before they escalate, unlike drones or weather stations, which often provide information only after problems have already progressed and plants experience severe damage. Early disease and crop stress detection ensures that farmers can act timely to maximize yields and optimize fertilizers and pesticides application. Cropler isn’t just about monitoring—it’s about driving actionable insights that make farming more efficient, sustainable, and profitable. Milestones and achievements 🏆 357 devices deployed
🏆 Revenue of $53k in 2024
🏆 AI disease & leaf damages integration
🏆 iOS & Android app release
🏆 10 dealer centers in multiple countries
🏆 16 countries (EU + USA) have chosen our devices
🏆 USA & EU issued certificates for our devices
🏆 EIT Food Test Farm winner Finalist of 2024
🏆 China Overseas Talents Innovation and Entrepreneurship Competition
🏆 Partnerships with key market players (Varda Field ID, AGRIVI, SAS Agro, Kalo, Basf (Xarvio), Bayer, Syngenta etc.) Lessons learned 1️⃣ This was quite a challenge in building AI tools that could accurately adapt to the vast complexities of agronomic data. Initial models developed were rather far from the precision that agronomists required, struggling with symptom and stage variability across different crops. However, we further invested in our AI capabilities, enhancing our models by adding extensive data sets and collaborations with agricultural experts. This achieved two things: it not only enhanced the precision of our AI but also made our tools an inseparable companion for any agronomist by providing them with second-to-none insights and recommendations. 2️⃣ Another challenge was to convince the highly practical and quite skeptical agronomists to adopt this high-tech tool called Cropler. So, we shifted strategy and focused our efforts on education and hands-on involvement, showing agronomists not only how but why our technology worked. 3️⃣ One of the most significant challenges of launching the international expansion of our solution was overcoming the many regulatory, environmental, and operational hurdles spanning across continents, which drastically changed from region to region. We learned flexibility and local partnerships. By aligning ourselves with local agronomic experts and adapting our technology to meet regional regulations and needs, the stage was set for global impact-development of preparations for a major rollout in Europe by 2025. 4️⃣ Last, but not least, was a technical and logistical challenge to scale our technology from pilot projects to widespread usage by agronomists. So, we invested in scalable architecture and robust systems, supporting high demand for real-time data processing and global access. Our platform is not a tool; it's a 24/7 digital agronomist turning that data into actionable and timely advice. What Startups of The Year means to us Entering into HackerNoon's Startups of the Year is not an opportunity for Cropler but rather a stage to revolutionize the agribusiness panorama. We come to introduce our pioneering AI agri-camera technology that's going to change the world and make agriculture more predictable and effective. This is a moment in the spotlight to galvanize an industry for the attention of global investors, tech enthusiasts, and potential partners who are committed to transforming agriculture through innovation as much as we are. To be present there is already a loud statement of the capability and vision that we have. More than that, we are not just in attendance; we are leading from the front in this charge into the future where technology and sustainability come together to remake the standard of agriculture around the world. The insights and feedback that we get here from this eclectic tech community are very critical; it fuels our drive toward the 2025 launch of our Global Analytics. For us, this is more than a competition; this is a launching pad toward growth, visibility, and impact. Join us while we lead the way to a more efficient, sustainable, and predictable agricultural future! Vote for us today: https://hackernoon.com/startups/europe/europe-warsaw-poland Hey Hackers! Cropler has been nominated in HackerNoon's Startups of The Year 2024 awards in Warsaw, Poland under the Emerging Tech category and others. Cropler Cropler HackerNoon's Startups of The Year 2024 HackerNoon's Startups of The Year 2024 HackerNoon's Startups of The Year 2024 Emerging Tech category and others. Emerging Tech Emerging Tech Please vote for us here: Please vote for us here: 1️⃣ https://hackernoon.com/startups/europe/europe-warsaw-poland https://hackernoon.com/startups/europe/europe-warsaw-poland https://hackernoon.com/startups/europe/europe-warsaw-poland 2️⃣ https://hackernoon.com/startups/industry/iot https://hackernoon.com/startups/industry/iot https://hackernoon.com/startups/industry/iot 3️⃣ https://hackernoon.com/startups/industry/hardware https://hackernoon.com/startups/industry/hardware 4️⃣ https://hackernoon.com/startups/industry/climatetech https://hackernoon.com/startups/industry/climatetech Meet Cropler Cropler is the world’s first real-time remote photo plant monitoring technology that is about to become an AI-powered digital agronomical assistant. Cropler uses smart AI-powered mobile agricameras to help farmers and agribusinesses identify the key phenophases, diseases, and inputs efficiently to perform necessary agricultural operations in a timely manner. Cropler is the world’s first real-time remote photo plant monitoring technology that is about to become an AI-powered digital agronomical assistant. Cropler uses smart AI-powered mobile agricameras to help farmers and agribusinesses identify the key phenophases, diseases, and inputs efficiently to perform necessary agricultural operations in a timely manner. Cropler Cropler agricameras The Winning Team 1️⃣ Nik Bedunkevich LinkedIn Founder & CEO With over 10 years in software and 7+ years in hardware development, Nik's leadership in precision agriculture, backed by his experience at OneSoil, where he co-founded and led hardware development, has been instrumental. His background in Radio Engineering and his ability to integrate data exchange and soil analysis sensors have enabled Cropler to push the boundaries of AI-driven agricultural solutions. 2️⃣ Marie Sinkova LinkedIn CMO Marie’s 6+ years in marketing, including her impactful work at MEDvidi and EduDo, have significantly boosted our brand’s visibility. Her ability to develop engaging content and manage successful teams has increased revenue and brand recognition while minimizing negative reviews. Her strategic approach in various marketing areas has laid a solid foundation for Cropler development. 3️⃣ Andrii Zbrytskyi LinkedIn BDM With over 17 years in the agritech area, Andri Zbrytskyi has witnessed firsthand the ins and outs of clients resulting in better client/corporate relations, sales and even more business development for the company. AI agriculture solutions have found strategic partnerships thanks to his efforts and these have increased expanding the market base. 4️⃣ Xenia Niaborskaya LinkedIn Head of Design Xenia’s experience of more than 4 years in UX/UI design has played a key role in ensuring that our user interfaces work as expected and look good as well. Her capability in business-centered design and the process of user-centered design has improved the usability of our platform tremendously. 5️⃣ Eugene Voronets LinkedIn Tech Adviser Eugene’s 10+ years in backend development, particularly his expertise in high-load systems, ensures the robustness and scalability of our technology. His work at OneSoil, building APIs and fostering collaborations with industry leaders like John Deere and AWS, has played a critical role in shaping Cropler’s technical infrastructure. 1️⃣ Nik Bedunkevich Nik Bedunkevich LinkedIn LinkedIn LinkedIn LinkedIn Founder & CEO Founder & CEO With over 10 years in software and 7+ years in hardware development, Nik's leadership in precision agriculture, backed by his experience at OneSoil, where he co-founded and led hardware development, has been instrumental. His background in Radio Engineering and his ability to integrate data exchange and soil analysis sensors have enabled Cropler to push the boundaries of AI-driven agricultural solutions. 2️⃣ Marie Sinkova Marie Sinkova LinkedIn LinkedIn LinkedIn LinkedIn CMO CMO Marie’s 6+ years in marketing, including her impactful work at MEDvidi and EduDo, have significantly boosted our brand’s visibility. Her ability to develop engaging content and manage successful teams has increased revenue and brand recognition while minimizing negative reviews. Her strategic approach in various marketing areas has laid a solid foundation for Cropler development. 3️⃣ Andrii Zbrytskyi Andrii Zbrytskyi LinkedIn LinkedIn LinkedIn LinkedIn BDM BDM With over 17 years in the agritech area, Andri Zbrytskyi has witnessed firsthand the ins and outs of clients resulting in better client/corporate relations, sales and even more business development for the company. AI agriculture solutions have found strategic partnerships thanks to his efforts and these have increased expanding the market base. 4️⃣ Xenia Niaborskaya Xenia Niaborskaya LinkedIn LinkedIn LinkedIn LinkedIn Head of Design Head of Design Xenia’s experience of more than 4 years in UX/UI design has played a key role in ensuring that our user interfaces work as expected and look good as well. Her capability in business-centered design and the process of user-centered design has improved the usability of our platform tremendously. 5️⃣ Eugene Voronets Eugene Voronets LinkedIn LinkedIn LinkedIn LinkedIn Tech Adviser Tech Adviser Eugene’s 10+ years in backend development, particularly his expertise in high-load systems, ensures the robustness and scalability of our technology. His work at OneSoil, building APIs and fostering collaborations with industry leaders like John Deere and AWS, has played a critical role in shaping Cropler’s technical infrastructure. What Cropler does differently We aim to revolutionize agriculture with Cropler's global plant monitoring technology, which uses AI-powered agricameras and remote sensing data to provide farmers, agronomists, and agribusinesses with predictive field insights and global yield analytics for optimized crop production. global plant monitoring technology, Cropler’s unique, AI-powered plant health data allows farmers and agribusinesses to address issues before they escalate, unlike drones or weather stations, which often provide information only after problems have already progressed and plants experience severe damage. Early disease and crop stress detection ensures that farmers can act timely to maximize yields and optimize fertilizers and pesticides application. Cropler isn’t just about monitoring—it’s about driving actionable insights that make farming more efficient, sustainable, and profitable. Cropler isn’t just about monitoring—it’s about driving actionable insights that make farming more efficient, sustainable, and profitable. Milestones and achievements 🏆 357 devices deployed 🏆 Revenue of $53k in 2024 🏆 AI disease & leaf damages integration 🏆 iOS & Android app release 🏆 10 dealer centers in multiple countries 🏆 16 countries (EU + USA) have chosen our devices 🏆 USA & EU issued certificates for our devices 🏆 EIT Food Test Farm winner Finalist of 2024 🏆 China Overseas Talents Innovation and Entrepreneurship Competition 🏆 Partnerships with key market players (Varda Field ID, AGRIVI, SAS Agro, Kalo, Basf (Xarvio), Bayer, Syngenta etc.) 🏆 357 devices deployed 🏆 Revenue of $53k in 2024 🏆 AI disease & leaf damages integration 🏆 iOS & Android app release 🏆 10 dealer centers in multiple countries 🏆 16 countries (EU + USA) have chosen our devices 🏆 USA & EU issued certificates for our devices 🏆 EIT Food Test Farm winner Finalist of 2024 🏆 China Overseas Talents Innovation and Entrepreneurship Competition 🏆 Partnerships with key market players (Varda Field ID, AGRIVI, SAS Agro, Kalo, Basf (Xarvio), Bayer, Syngenta etc.) Lessons learned 1️⃣ This was quite a challenge in building AI tools that could accurately adapt to the vast complexities of agronomic data. Initial models developed were rather far from the precision that agronomists required, struggling with symptom and stage variability across different crops. However, we further invested in our AI capabilities, enhancing our models by adding extensive data sets and collaborations with agricultural experts. This achieved two things: it not only enhanced the precision of our AI but also made our tools an inseparable companion for any agronomist by providing them with second-to-none insights and recommendations. 2️⃣ Another challenge was to convince the highly practical and quite skeptical agronomists to adopt this high-tech tool called Cropler. So, we shifted strategy and focused our efforts on education and hands-on involvement, showing agronomists not only how but why our technology worked. agronomists 3️⃣ One of the most significant challenges of launching the international expansion of our solution was overcoming the many regulatory, environmental, and operational hurdles spanning across continents, which drastically changed from region to region. We learned flexibility and local partnerships. By aligning ourselves with local agronomic experts and adapting our technology to meet regional regulations and needs, the stage was set for global impact-development of preparations for a major rollout in Europe by 2025. global impact-development 4️⃣ Last, but not least, was a technical and logistical challenge to scale our technology from pilot projects to widespread usage by agronomists. So, we invested in scalable architecture and robust systems, supporting high demand for real-time data processing and global access. Our platform is not a tool; it's a 24/7 digital agronomist turning that data into actionable and timely advice. Our platform is not a tool; it's a 24/7 digital agronomist turning that data into actionable and timely advice. Our platform is not a tool; it's a 24/7 digital agronomist turning that data into actionable and timely advice. What Startups of The Year means to us Entering into HackerNoon's Startups of the Year is not an opportunity for Cropler but rather a stage to revolutionize the agribusiness panorama. We come to introduce our pioneering AI agri-camera technology that's going to change the world and make agriculture more predictable and effective. This is a moment in the spotlight to galvanize an industry for the attention of global investors, tech enthusiasts, and potential partners who are committed to transforming agriculture through innovation as much as we are. revolutionize the agribusiness To be present there is already a loud statement of the capability and vision that we have. To be present there is already a loud statement of the capability and vision that we have. More than that, we are not just in attendance; we are leading from the front in this charge into the future where technology and sustainability come together to remake the standard of agriculture around the world. The insights and feedback that we get here from this eclectic tech community are very critical; it fuels our drive toward the 2025 launch of our Global Analytics. For us, this is more than a competition; this is a launching pad toward growth, visibility, and impact. For us, this is more than a competition; this is a launching pad toward growth, visibility, and impact. Join us while we lead the way to a more efficient, sustainable, and predictable agricultural future! Vote for us today: https://hackernoon.com/startups/europe/europe-warsaw-poland Vote for us today: https://hackernoon.com/startups/europe/europe-warsaw-poland Vote for us today: https://hackernoon.com/startups/europe/europe-warsaw-poland

AI-Driven Solutions for Modern Agritech: Interview with SOTY 2024 Nominee, Cropler

Authors:
(1) Divyanshu Kumar, Enkrypt AI;
(2) Anurakt Kumar, Enkrypt AI;
(3) Sahil Agarwa, Enkrypt AI;
(4) Prashanth Harshangi, Enkrypt AI. Table of Links Abstract and 1 Introduction 2 Problem Formulation and Experiments 3 Experiment Set-up & Results 4 Conclusion and References A. Appendix 4 CONCLUSION Our work investigates the LLM’s safety against Jailbreak attempts. We have demonstrated how fine-tuned and quantized models are vulnerable to jailbreak attempts and stress the importance of using external guardrails to reduce this risk. Fine-tuning or quantizing model weights alters the risk profile of LLMs, potentially undermining the safety alignment established through RLHF. This could result from catastrophic forgetting, where LLMs lose memory of safety protocols, or the finetuning process shifting the model’s focus to new topics at the expense of existing safety measures. The lack of safety measures in these fine-tuned and quantized models is concerning, highlighting the need to incorporate safety protocols during the fine-tuning process. We propose using these tests as part of a CI/CD stress test before deploying the model. The effectiveness of guardrails in preventing jailbreaking highlights the importance of integrating them with safety practices in AI development. This approach not only enhances AI models but also establishes a new standard for responsible AI development. By ensuring that AI advancements prioritize innovation and safety, we promote ethical AI deployment, safeguarding against potential misuse and fostering a secure digital future. REFERENCES Zifan Wang Andy Zou. AdvBench Dataset, July 2023. URL https://github.com/ llm-attacks/llm-attacks/tree/main/data. Patrick Chao, Alexander Robey, Edgar Dobriban, Hamed Hassani, George J. Pappas, and Eric Wong. Jailbreaking Black Box Large Language Models in Twenty Queries. arXiv, October 2023. doi: 10.48550/arXiv.2310.08419. Edoardo Debenedetti, Giorgio Severi, Nicholas Carlini, Christopher A. Choquette-Choo, Matthew Jagielski, Milad Nasr, Eric Wallace, and Florian Tramer. Privacy Side Channels in Machine ` Learning Systems. arXiv, September 2023. doi: 10.48550/arXiv.2309.05610. Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, and Luke Zettlemoyer. QLoRA: Efficient Finetuning of Quantized LLMs. arXiv, May 2023. doi: 10.48550/arXiv.2305.14314. Micah Gorsline, James Smith, and Cory Merkel. On the Adversarial Robustness of Quantized Neural Networks. arXiv, May 2021. doi: 10.1145/3453688.3461755. Kai Greshake, Sahar Abdelnabi, Shailesh Mishra, Christoph Endres, Thorsten Holz, and Mario Fritz. Not what you’ve signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection. arXiv, February 2023. doi: 10.48550/arXiv.2302.12173. Bing He, Mustaque Ahamad, and Srijan Kumar. PETGEN: Personalized Text Generation Attack on Deep Sequence Embedding-based Classification Models. In KDD ’21: Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pp. 575–584. Association for Computing Machinery, New York, NY, USA, August 2021. ISBN 978-1-45038332-5. doi: 10.1145/3447548.3467390. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. LoRA: Low-Rank Adaptation of Large Language Models. arXiv, June 2021. doi: 10.48550/arXiv.2106.09685. Shuhei Kashiwamura, Ayaka Sakata, and Masaaki Imaizumi. Effect of Weight Quantization on Learning Models by Typical Case Analysis. arXiv, January 2024. doi: 10.48550/arXiv.2401. 17269. Siwon Kim, Sangdoo Yun, Hwaran Lee, Martin Gubri, Sungroh Yoon, and Seong Joon Oh. ProPILE: Probing Privacy Leakage in Large Language Models. arXiv, July 2023. doi: 10.48550/arXiv. 2307.01881. Aounon Kumar, Chirag Agarwal, Suraj Srinivas, Aaron Jiaxun Li, Soheil Feizi, and Himabindu Lakkaraju. Certifying LLM Safety against Adversarial Prompting. arXiv, September 2023. doi: 10.48550/arXiv.2309.02705. Thai Le, Suhang Wang, and Dongwon Lee. MALCOM: Generating Malicious Comments to Attack Neural Fake News Detection Models. IEEE Computer Society, November 2020. ISBN 978-1- 7281-8316-9. doi: 10.1109/ICDM50108.2020.00037. Yi Liu, Gelei Deng, Zhengzi Xu, Yuekang Li, Yaowen Zheng, Ying Zhang, Lida Zhao, Tianwei Zhang, and Yang Liu. Jailbreaking ChatGPT via Prompt Engineering: An Empirical Study. arXiv, May 2023. doi: 10.48550/arXiv.2305.13860. Anay Mehrotra, Manolis Zampetakis, Paul Kassianik, Blaine Nelson, Hyrum Anderson, Yaron Singer, and Amin Karbasi. Tree of Attacks: Jailbreaking Black-Box LLMs Automatically. arXiv, December 2023. doi: 10.48550/arXiv.2312.02119. Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul Christiano, Jan Leike, and Ryan Lowe. Training language models to follow instructions with human feedback. arXiv, March 2022. doi: 10.48550/arXiv.2203.02155. Xiangyu Qi, Yi Zeng, Tinghao Xie, Pin-Yu Chen, Ruoxi Jia, Prateek Mittal, and Peter Henderson. Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To! arXiv, October 2023. doi: 10.48550/arXiv.2310.03693. Traian Rebedea, Razvan Dinu, Makesh Narsimhan Sreedhar, Christopher Parisien, and Jonathan Cohen. NeMo Guardrails: A Toolkit for Controllable and Safe LLM Applications with Programmable Rails. ACL Anthology, pp. 431–445, December 2023. doi: 10.18653/v1/2023. emnlp-demo.40. Alexander Wei, Nika Haghtalab, and Jacob Steinhardt. Jailbroken: How Does LLM Safety Training Fail? arXiv, July 2023. doi: 10.48550/arXiv.2307.02483. Martin Weyssow, Xin Zhou, Kisub Kim, David Lo, and Houari Sahraoui. Exploring ParameterEfficient Fine-Tuning Techniques for Code Generation with Large Language Models. arXiv, August 2023. doi: 10.48550/arXiv.2308.10462. Yisong Xiao, Aishan Liu, Tianyuan Zhang, Haotong Qin, Jinyang Guo, and Xianglong Liu. RobustMQ: Benchmarking Robustness of Quantized Models. arXiv, August 2023. doi: 10.48550/ arXiv.2308.02350. Andy Zhou, Bo Li, and Haohan Wang. Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks. arXiv, January 2024. doi: 10.48550/arXiv.2401.17263. Sicheng Zhu, Ruiyi Zhang, Bang An, Gang Wu, Joe Barrow, Zichao Wang, Furong Huang, Ani Nenkova, and Tong Sun. AutoDAN: Interpretable Gradient-Based Adversarial Attacks on Large Language Models. arXiv, October 2023. doi: 10.48550/arXiv.2310.15140. Andy Zou, Zifan Wang, Nicholas Carlini, Milad Nasr, J. Zico Kolter, and Matt Fredrikson. Universal and Transferable Adversarial Attacks on Aligned Language Models. arXiv, July 2023. doi: 10.48550/arXiv.2307.15043. This paper is available on arxiv under CC BY-NC-SA 4.0 DEED license. Authors: (1) Divyanshu Kumar, Enkrypt AI; (2) Anurakt Kumar, Enkrypt AI; (3) Sahil Agarwa, Enkrypt AI; (4) Prashanth Harshangi, Enkrypt AI. Authors: Authors: (1) Divyanshu Kumar, Enkrypt AI; (2) Anurakt Kumar, Enkrypt AI; (3) Sahil Agarwa, Enkrypt AI; (4) Prashanth Harshangi, Enkrypt AI. Table of Links Abstract and 1 Introduction Abstract and 1 Introduction 2 Problem Formulation and Experiments 2 Problem Formulation and Experiments 3 Experiment Set-up & Results 3 Experiment Set-up & Results 4 Conclusion and References 4 Conclusion and References A. Appendix A. Appendix 4 CONCLUSION Our work investigates the LLM’s safety against Jailbreak attempts. We have demonstrated how fine-tuned and quantized models are vulnerable to jailbreak attempts and stress the importance of using external guardrails to reduce this risk. Fine-tuning or quantizing model weights alters the risk profile of LLMs, potentially undermining the safety alignment established through RLHF. This could result from catastrophic forgetting, where LLMs lose memory of safety protocols, or the finetuning process shifting the model’s focus to new topics at the expense of existing safety measures. The lack of safety measures in these fine-tuned and quantized models is concerning, highlighting the need to incorporate safety protocols during the fine-tuning process. We propose using these tests as part of a CI/CD stress test before deploying the model. The effectiveness of guardrails in preventing jailbreaking highlights the importance of integrating them with safety practices in AI development. This approach not only enhances AI models but also establishes a new standard for responsible AI development. By ensuring that AI advancements prioritize innovation and safety, we promote ethical AI deployment, safeguarding against potential misuse and fostering a secure digital future. REFERENCES Zifan Wang Andy Zou. AdvBench Dataset, July 2023. URL https://github.com/ llm-attacks/llm-attacks/tree/main/data. Patrick Chao, Alexander Robey, Edgar Dobriban, Hamed Hassani, George J. Pappas, and Eric Wong. Jailbreaking Black Box Large Language Models in Twenty Queries. arXiv, October 2023. doi: 10.48550/arXiv.2310.08419. Edoardo Debenedetti, Giorgio Severi, Nicholas Carlini, Christopher A. Choquette-Choo, Matthew Jagielski, Milad Nasr, Eric Wallace, and Florian Tramer. Privacy Side Channels in Machine ` Learning Systems. arXiv, September 2023. doi: 10.48550/arXiv.2309.05610. Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, and Luke Zettlemoyer. QLoRA: Efficient Finetuning of Quantized LLMs. arXiv, May 2023. doi: 10.48550/arXiv.2305.14314. Micah Gorsline, James Smith, and Cory Merkel. On the Adversarial Robustness of Quantized Neural Networks. arXiv, May 2021. doi: 10.1145/3453688.3461755. Kai Greshake, Sahar Abdelnabi, Shailesh Mishra, Christoph Endres, Thorsten Holz, and Mario Fritz. Not what you’ve signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection. arXiv, February 2023. doi: 10.48550/arXiv.2302.12173. Bing He, Mustaque Ahamad, and Srijan Kumar. PETGEN: Personalized Text Generation Attack on Deep Sequence Embedding-based Classification Models. In KDD ’21: Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pp. 575–584. Association for Computing Machinery, New York, NY, USA, August 2021. ISBN 978-1-45038332-5. doi: 10.1145/3447548.3467390. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. LoRA: Low-Rank Adaptation of Large Language Models. arXiv, June 2021. doi: 10.48550/arXiv.2106.09685. Shuhei Kashiwamura, Ayaka Sakata, and Masaaki Imaizumi. Effect of Weight Quantization on Learning Models by Typical Case Analysis. arXiv, January 2024. doi: 10.48550/arXiv.2401. 17269. Siwon Kim, Sangdoo Yun, Hwaran Lee, Martin Gubri, Sungroh Yoon, and Seong Joon Oh. ProPILE: Probing Privacy Leakage in Large Language Models. arXiv, July 2023. doi: 10.48550/arXiv. 2307.01881. Aounon Kumar, Chirag Agarwal, Suraj Srinivas, Aaron Jiaxun Li, Soheil Feizi, and Himabindu Lakkaraju. Certifying LLM Safety against Adversarial Prompting. arXiv, September 2023. doi: 10.48550/arXiv.2309.02705. Thai Le, Suhang Wang, and Dongwon Lee. MALCOM: Generating Malicious Comments to Attack Neural Fake News Detection Models. IEEE Computer Society, November 2020. ISBN 978-1- 7281-8316-9. doi: 10.1109/ICDM50108.2020.00037. Yi Liu, Gelei Deng, Zhengzi Xu, Yuekang Li, Yaowen Zheng, Ying Zhang, Lida Zhao, Tianwei Zhang, and Yang Liu. Jailbreaking ChatGPT via Prompt Engineering: An Empirical Study. arXiv, May 2023. doi: 10.48550/arXiv.2305.13860. Anay Mehrotra, Manolis Zampetakis, Paul Kassianik, Blaine Nelson, Hyrum Anderson, Yaron Singer, and Amin Karbasi. Tree of Attacks: Jailbreaking Black-Box LLMs Automatically. arXiv, December 2023. doi: 10.48550/arXiv.2312.02119. Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul Christiano, Jan Leike, and Ryan Lowe. Training language models to follow instructions with human feedback. arXiv, March 2022. doi: 10.48550/arXiv.2203.02155. Xiangyu Qi, Yi Zeng, Tinghao Xie, Pin-Yu Chen, Ruoxi Jia, Prateek Mittal, and Peter Henderson. Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To! arXiv, October 2023. doi: 10.48550/arXiv.2310.03693. Traian Rebedea, Razvan Dinu, Makesh Narsimhan Sreedhar, Christopher Parisien, and Jonathan Cohen. NeMo Guardrails: A Toolkit for Controllable and Safe LLM Applications with Programmable Rails. ACL Anthology, pp. 431–445, December 2023. doi: 10.18653/v1/2023. emnlp-demo.40. Alexander Wei, Nika Haghtalab, and Jacob Steinhardt. Jailbroken: How Does LLM Safety Training Fail? arXiv, July 2023. doi: 10.48550/arXiv.2307.02483. Martin Weyssow, Xin Zhou, Kisub Kim, David Lo, and Houari Sahraoui. Exploring ParameterEfficient Fine-Tuning Techniques for Code Generation with Large Language Models. arXiv, August 2023. doi: 10.48550/arXiv.2308.10462. Yisong Xiao, Aishan Liu, Tianyuan Zhang, Haotong Qin, Jinyang Guo, and Xianglong Liu. RobustMQ: Benchmarking Robustness of Quantized Models. arXiv, August 2023. doi: 10.48550/ arXiv.2308.02350. Andy Zhou, Bo Li, and Haohan Wang. Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks. arXiv, January 2024. doi: 10.48550/arXiv.2401.17263. Sicheng Zhu, Ruiyi Zhang, Bang An, Gang Wu, Joe Barrow, Zichao Wang, Furong Huang, Ani Nenkova, and Tong Sun. AutoDAN: Interpretable Gradient-Based Adversarial Attacks on Large Language Models. arXiv, October 2023. doi: 10.48550/arXiv.2310.15140. Andy Zou, Zifan Wang, Nicholas Carlini, Milad Nasr, J. Zico Kolter, and Matt Fredrikson. Universal and Transferable Adversarial Attacks on Aligned Language Models. arXiv, July 2023. doi: 10.48550/arXiv.2307.15043. This paper is available on arxiv under CC BY-NC-SA 4.0 DEED license. This paper is available on arxiv under CC BY-NC-SA 4.0 DEED license. available on arxiv

Increased LLM Vulnerabilities from Fine-tuning and Quantization: Conclusion and References

Apparate: Early-Exit Models for ML Latency and Throughput Optimization - Conclusion, References

Constant innovation drives the evolution of Web3, aiming to improve user experiences and developer workflows. Account Abstraction (AA) has been gathering a lot of buzz recently, and it is a concept designed to simplify interactions within the Web3 ecosystem. Today, we will dive into the realm of Modular Accounts which are the next step in the AA approach. Understanding Modular Accounts Modular Accounts are a groundbreaking concept within account abstraction that seeks to offer greater customization for both users and developers. At its core, Modular Accounts are structured as composable plugins or modules, where each module represents a distinct smart contract extending the functionality of a core smart account. This modular architecture allows tailored customization, empowering developers to focus on specific features rather than rebuilding entire account infrastructures. Modular Accounts Standardization: ERC-6900 vs. ERC-7579 There are initiatives such as ERC-6900 and ERC-7579 that aim to enhance interoperability and standardization of Modular Accounts. These standards define how modules interact with each other and with smart accounts across different providers and platforms. ERC-6900, introduced in April 2023, establishes comprehensive standards for smart account and module interfaces. While robust, its complexity and opinionated security measures may pose challenges for developers seeking flexibility. Even though the authors of ERC-7579 support the objective of ERC-6900 to standardize modular accounts, they believe that ERC-7579 adopts a minimally prescriptive approach, prioritizing interoperability, and not discouraging innovation. This approach allows developers to make decisions around complexity and security, fostering a competitive ecosystem of modular accounts and modules. Etherspot Has Launched Its Modular SDK Etherspot is proud to introduce Modular Accounts that are compliant with the ERC-7579 standard, empowering developers leveraging Etherspot’s Account Abstraction infrastructure with customization capabilities for smart accounts. Our smart contracts have been audited by KALOS, a reputable Web3 auditing company established in 2018, ensuring security and reliability for our users. Developers can significantly enhance the security and user experience of their dApps by integrating various modules. For instance, social recovery modules empower users to regain account access through a trusted device or friends, in case access has been lost. Another feature is a module allowing dApps to spend tokens on behalf of users within a predefined limit, streamlining transactions and reducing manual interventions. Additionally, session key modules enable users to create a session key that pre-approves transactions within a dApp for a specified duration, under set limits and rules, offering both convenience and security. Etherspot’s session key module has been audited by Shieldify, ensuring a high standard of trust and reliability. The list of existing modules can be found here. Developers can install any module with Etherspot’s Modular SDK. By adopting ERC-7579 compliance, Etherspot is committed to driving innovation and interoperability in the Web3 ecosystem, enabling developers to build scalable and customizable decentralized applications with confidence. ⚙️To get started with Etherspot Modular SDK, you can either install the packages yourself: npm i @etherspot/modular-sdk Find more info in our documentation here. Alternatively, you can leverage our robust React library — TransactionKit, which already supports modular accounts: https://etherspot.fyi/transaction-kit/intro. Modular Store For Plug-n-play Solutions ERC-7579 opens up an exciting world of modular components and user-friendly tools for developers. This means easier access to reusable modules, allowing developers to focus on creating unique features that set their products apart while speeding up their development process. Rhinestone is building a modular marketplace where developers can discover suitable modules and seamlessly integrate them into Smart Accounts, as well as create and share their own modules. They’ve started with some essential modules that anyone can use. It’s still under development, follow Rhinestone on X to stay up to date. At the same time, the Etherspot team has been working on the Modular Store to be released on the PillarX (PX) platform. The PX Modular Store is beneficial to both developers and users. It will allow users to install modules as Add-Ons to their smart accounts, while dApps can leverage modules to deliver an even smoother account abstraction experience for users within the PX ecosystem. Follow PillarX on X to keep up with the latest updates. As the Web3 landscape continues to evolve, innovations like Modular Accounts pave the way for a more accessible, flexible, and secure decentralized ecosystem. The Etherspot Modular SDK’s compliance with the ERC-7579 promises a simplified path toward building scalable and interoperable dApps for developers interested in exploring the potential of Modular Accounts within Etherspot’s infrastructure Resources: Introducing: ERC-7579
WTF is Modular Account Abstraction
Awesome Modular Accounts Follow Etherspot: Website | X | Discord | Telegram | GitHub Constant innovation drives the evolution of Web3, aiming to improve user experiences and developer workflows. Account Abstraction (AA) has been gathering a lot of buzz recently, and it is a concept designed to simplify interactions within the Web3 ecosystem. Today, we will dive into the realm of Modular Accounts which are the next step in the AA approach. Understanding Modular Accounts Modular Accounts are a groundbreaking concept within account abstraction that seeks to offer greater customization for both users and developers. At its core, Modular Accounts are structured as composable plugins or modules, where each module represents a distinct smart contract extending the functionality of a core smart account. This modular architecture allows tailored customization, empowering developers to focus on specific features rather than rebuilding entire account infrastructures. Modular Accounts Standardization: ERC-6900 vs. ERC-7579 There are initiatives such as ERC-6900 and ERC-7579 that aim to enhance interoperability and standardization of Modular Accounts. These standards define how modules interact with each other and with smart accounts across different providers and platforms. ERC-6900 , introduced in April 2023, establishes comprehensive standards for smart account and module interfaces. While robust, its complexity and opinionated security measures may pose challenges for developers seeking flexibility. ERC-6900 ERC-6900 Even though the authors of ERC-7579 support the objective of ERC-6900 to standardize modular accounts, they believe that ERC-7579 adopts a minimally prescriptive approach, prioritizing interoperability, and not discouraging innovation. This approach allows developers to make decisions around complexity and security, fostering a competitive ecosystem of modular accounts and modules. ERC-7579 ERC-7579 Etherspot Has Launched Its Modular SDK Etherspot is proud to introduce Modular Accounts that are compliant with the ERC-7579 standard, empowering developers leveraging Etherspot’s Account Abstraction infrastructure with customization capabilities for smart accounts. Our smart contracts have been audited by KALOS , a reputable Web3 auditing company established in 2018, ensuring security and reliability for our users. have been audited have been audited KALOS KALOS Developers can significantly enhance the security and user experience of their dApps by integrating various modules. For instance, social recovery modules empower users to regain account access through a trusted device or friends, in case access has been lost. Another feature is a module allowing dApps to spend tokens on behalf of users within a predefined limit, streamlining transactions and reducing manual interventions. Additionally, session key modules enable users to create a session key that pre-approves transactions within a dApp for a specified duration, under set limits and rules, offering both convenience and security. Etherspot’s session key module has been audited by Shieldify , ensuring a high standard of trust and reliability. Etherspot’s session key module Etherspot’s session key module audited by Shieldify audited by Shieldify The list of existing modules can be found here . Developers can install any module with Etherspot’s Modular SDK. here here By adopting ERC-7579 compliance, Etherspot is committed to driving innovation and interoperability in the Web3 ecosystem, enabling developers to build scalable and customizable decentralized applications with confidence. ⚙️To get started with Etherspot Modular SDK, you can either install the packages yourself: npm i @etherspot/modular-sdk npm i @etherspot/modular-sdk Find more info in our documentation here . here here Alternatively, you can leverage our robust React library — TransactionKit , which already supports modular accounts: https://etherspot.fyi/transaction-kit/intro . TransactionKit https://etherspot.fyi/transaction-kit/intro https://etherspot.fyi/transaction-kit/intro Modular Store For Plug-n-play Solutions ERC-7579 opens up an exciting world of modular components and user-friendly tools for developers. This means easier access to reusable modules, allowing developers to focus on creating unique features that set their products apart while speeding up their development process. Rhinestone is building a modular marketplace where developers can discover suitable modules and seamlessly integrate them into Smart Accounts, as well as create and share their own modules. They’ve started with some essential modules that anyone can use. It’s still under development, follow Rhinestone on X to stay up to date. Rhinestone Rhinestone modular marketplace X X At the same time, the Etherspot team has been working on the Modular Store to be released on the PillarX (PX) platform . The PX Modular Store is beneficial to both developers and users. PillarX (PX) platform PillarX (PX) platform It will allow users to install modules as Add-Ons to their smart accounts, while dApps can leverage modules to deliver an even smoother account abstraction experience for users within the PX ecosystem. Follow PillarX on X to keep up with the latest updates. PillarX on X PillarX on X As the Web3 landscape continues to evolve, innovations like Modular Accounts pave the way for a more accessible, flexible, and secure decentralized ecosystem. The Etherspot Modular SDK’s compliance with the ERC-7579 promises a simplified path toward building scalable and interoperable dApps for developers interested in exploring the potential of Modular Accounts within Etherspot’s infrastructure Resources: Resources: Introducing: ERC-7579 WTF is Modular Account Abstraction Awesome Modular Accounts Introducing: ERC-7579 Introducing: ERC-7579 Introducing: ERC-7579 WTF is Modular Account Abstraction WTF is Modular Account Abstraction WTF is Modular Account Abstraction Awesome Modular Accounts Awesome Modular Accounts Awesome Modular Accounts Follow Etherspot: Follow Etherspot: Website | X | Discord | Telegram | GitHub Website X Discord Telegram GitHub

Etherspot Launches ERC-7579 Modular SDK: What You Need to Know

Etherspot ERC-7579 মডুলার SDK চালু করেছে: আপনার যা জানা দরকার

Etherspot bringt modulares SDK ERC-7579 auf den Markt: Was Sie wissen müssen

Etherspot lanza el SDK modular ERC-7579: lo que necesita saber

Etherspot lance le SDK modulaire ERC-7579 : ce que vous devez savoir

इथरस्पॉट ने ERC-7579 मॉड्यूलर SDK लॉन्च किया: आपको क्या जानना चाहिए

Etherspot が ERC-7579 モジュラー SDK をリリース: 知っておくべきこと

Etherspot이 ERC-7579 모듈형 SDK를 출시: 알아야 할 사항

Etherspot lança o SDK modular ERC-7579: o que você precisa saber

Etherspot запускает модульный SDK ERC-7579: что вам нужно знать

Etherspot ERC-7579 Modüler SDK'sını Piyasaya Sürüyor: Bilmeniz Gerekenler

Etherspot ra mắt SDK mô-đun ERC-7579: Những điều bạn cần biết

Etherspot 推出 ERC-7579 模块化 SDK：您需要了解的内容

Stories, News & Mentions about KALTO

Kalto’s Stories on HackerNoon