For Job Seekers: allwhere has been informed of fraudulent job postings being shared for positions at our company. Please take note that allwhere will never ask job seekers for payment as part of the application or interview process. If you become suspicious of a potential scam, please report the job posting right away. Thank you. allwhere is a people-first company that empowers teams to stay productive and engaged — wherever they may be. We work with companies to implement holistic remote work solutions, manage the full lifecycle of equipment, and provide world-class onboarding, engagement, retention, and wellness programs, all while lowering costs.

allwhere

Hi everyone! My name is Michil Androsov and this is the second article in my series on the history of , where I offer an inside look at the creation of our unicorn company from the perspective of a simple developer. The first article is available . In it, I talked about how I joined the company, how we expanded the development team, and how we went international. inDrive here This article will cover the middle period of my career at inDrive—from about 2016 to 2020. At the time, the company was beginning to stand on its own two feet and rapidly develop in all kinds of new directions, which I'll discuss in more detail below. The Business After our first successful launches outside of Yakutsk, we were finally starting to recognize the power of our business idea. We began to wonder, "What if we really manage to take over the planet?" After all, living far up north, , you always feel quite separate from the rest of the world. It seems like everything's happening so far away, which actually gives us a pretty unique perspective. And at the time, all of our work with inDrive felt almost like a game to me. deep in Siberia Over this period, the company began to launch extensively. When we'd start operations in a new country, we might launch dozens of cities within a single month. We expanded beyond Russia and launched in Kazakhstan, where—despite a variety of obstacles—we quickly became the country's №1 ride-hailing service. To this day, Kazakhstan remains one of our top markets. When launching in a new city, we adopted a strategy of " ". As our CEO would say, if a city was ready for our service, it should be able to ignite from the smallest of sparks. So, if the grass in one city wasn't dry enough, we wouldn't try to soak it in gasoline. Instead, we'd just move on to the next one. This helped us expand quickly without spending too much money. burning dry grass In fact, a number of cities simply launched on their own. People would hear about this great new service, called inDrive, in a neighboring city. They'd download the app for themselves, and the city would launch without us. There were even a few self-launched cities that took off without any investment on our part. Of course, not all launches were a success. I remember our first back in 2018. We made all the preparations, recruited drivers, and advertised our service. At first, everything seemed to be working out, and our numbers were growing day by day. USA launch We were pleasantly surprised. Was it really that easy to launch in one of the world's largest markets? After all, if you can make it in the US, you can make it anywhere—at least, that's what I thought. But our joy quickly became disappointment: We were faced with a huge industry of fraudsters cashing out stolen credit cards and using them on our service. So, we decided to scrap our American plans and wait for better days—which finally came in 2023. We launched for the second time in the USA, with much better results. This launch is still ongoing. We are currently operating in Miami, and will soon expand to new cities. Looking back, we were being simple and naive. However, we continued to learn from our mistakes and push forward. The Product and Technology We've always believed that our service should be simple and fail-proof, like a . Of course, this has led to accusations that the app isn't "pretty" enough, despite being packed with features. Nokia 3310 In many countries, we had to adapt to local legislation and user needs. That’s why the inDrive app in Almaty might look completely different from the one in New York. Nevertheless, we have been working on making the app "prettier" in all countries over the past years, and I personally think it looks quite modern now. Outside of Russia, inDrive gained a range of new modules, starting with intercity rides and freight delivery, which are now available in countries across the globe. In certain markets, such as Kazakhstan, they are even more popular than our taxi vertical. During this time, our approach to product planning was simple: At the end of every quarter, our CEO would arrive back from his trips to our launch cities, open his laptop, and explain our goals for the upcoming period. Usually, these came from driver and passenger requests. We could have tasks of all different sizes, from introducing card payments to changing the color of the order button. Our CEO played (and continues to play) a crucial role in the development of the company. While he is both an excellent motivator and strategist, he’s also a skilled manager who understands how the product works. Even now, he personally flies to new launch countries, studies the market, conducts customer development, interacts extensively with users, and gathers feedback on the service. On the development side, we quickly realized that our existing code would not be able to support our rapid business growth. Soon, the app would start to lag and decline in quality. So, we decided to rewrite the more "complex" parts of the code. In addition to the PHP monolith, we eventually got ourselves a— —Golang monolith. The entirety of our infrastructure lay on the shoulders of two guys who knew all the inner workings of our enormous machine. If the service ever crashed, they were prepared to turn on their laptops and fix it at any time, day or night. ba-dum-tss We also got our first QA specialist for full-scale testing. This made our lives incredibly easier. Before our QA, we had been spending a lot of time and effort testing new functionality and updates by ourselves, and checking how they interacted with existing versions. Little by little, mobile developers began to join our team, albeit with some difficulty. After all, you needed Apple products to get into iOS development, and those could be difficult to find in Yakutsk. This meant that we struggled to recruit local iOS developers, and our iOS team would sometimes lag behind our Android team when it came to releasing new features. I remember one time in Kazakhstan, when our app was blocked at the government level. I won't get into the details, but to reverse the block, we had to introduce a number of different features, including user instructions, automatic address changes, proxies, and more. In the end, we restored our capabilities in the country and officially reversed the block after a few years. I remember writing to the CTO of a large CIS news site that had also been blocked in Kazakhstan, and asking how he had managed to keep business going. Many of our features were developed through a series of trials and errors—or stumbled upon by chance—although they might seem completely obvious in hindsight. Take our call-free ride-hailing system, for example. If you haven't read the , I'll quickly summarize it here. first article Previously, when an inDrive passenger would create their order, a driver would see their request in the orders feed and call them right away. And this wouldn't be just one driver either; these orders were being sent to anyone nearby. Of course, this led to situations where multiple drivers would see an order and try to call the passenger at once. Typically, the first person to contact the passenger would get the order. Here's where the problem came in: Drivers would click on a ride without reading the order details, and then all try to call the passenger at the same time. So, we created something called a "buffer." When drivers press the order button, there's a short delay, which gives us time to collect all potential offers. This way, drivers don't have to rush, and they have time to read the order details in full. Then, based on the drivers' ratings, the system selects which of them will be able to call the passenger first. This was a significant improvement to our old version. Although the call-free system is now used by practically every ride-hailing app, we actually came up with ours by chance. In 2016, we launched in Yekaterinburg and introduced the "Guardians," who were specially selected drivers that could accept orders that typical drivers couldn't. We would send them special push notifications about exclusive orders. However, we soon realized that these push notifications weren't working very well because they were easy to miss. And that's when we remembered one of our existing features, called "personal orders," which allowed the passenger to simply select the closest driver on the map and send the order straight to them. We decided to use this feature when sending orders to our "Guardians." It took off immediately and worked so well that we then decided to expand it to all drivers. Of course, the personal orders feature might seem obvious now, but we discovered it through a series of trials and errors. The Team Up until 2019, our team was fully made up of Yakutsk locals. At the time, we wanted to help our homeland grow — and the local people along with it. inDrive had always held a strong mission and values. These values resonated with each employee, and we strived to uphold them. This was a period of powerful drive and energy. Every day, I would wake up in a good mood, knowing that my friends and good coffee were waiting for me in the office. That and plenty of interesting work, of course. inDrive was known for being a social company. Our team was young and carefree, which meant we worked hard and played hard. Every year, we'd take a company trip abroad, whether to Egypt, Thailand, Kazakhstan, or somewhere else entirely. Imagine an entire plane of colleagues escaping the –50-degree Celsius weather in Yakutsk to visit sunny +35-degree Thailand! It was great at the time, but unfortunately, these big company trips aren't possible anymore with a workforce of several thousand employees. By 2019, we hit a hiring roadblock. Business was booming, and any professionals in Yakutsk capable of handling the tasks and workload were already working for us. So, we decided to open a development office in Moscow. As the head of the iOS team, I was charged with opening the Moscow branch. We had a lot of flexibility during this time, and employees were free to contribute to the company in any way they saw fit, as long as it benefited the business. I entered a new phase in my life. I moved to Moscow and started the process of opening the new office, from choosing a business center to installing special lights with the company logo. Now that a few years have passed, I can look back at that time fondly. But back then, I would often think: "Why did I do this to myself? Why'd I take on all this renovation work?" Our main problems had to do with the company in charge of the renovation. At one point, they were so behind schedule that I even had to help them install the electrical wiring in the workspaces. I remember one specific evening, when we stayed late to assemble furniture. I could see that my colleagues were tired and depressed, so to lighten the mood, I asked " " At the time (and to this day), Belorusskaya was one of the most expensive areas in Moscow, like Times Square in New York. It was also home to a number of impressive IT offices. Of course, we just laughed and got back to work. Why the long faces? Cheer up! Next year, we're opening an office on Belorusskaya! By the end of 2019, we had our first developers at the Moscow office. Sometimes, it would seem like they were from another planet. During their interviews, they'd say things like, "We work in sprints. I finished all my tasks, then came here" (and it'd only be Thursday afternoon). Or they'd say, "We work three days from home and two days from the office." That was something completely new for me. inDrive was entirely office-based at the time, and everyone was expected to contribute fully. We weren't working with any kind of Kanbans, scrums, story points, or sprints. Conclusion We entered 2020 with a lot of optimism. Our new development office in Moscow had opened, and we were working with a great team. I was living in the capital, business was growing, and investments were on their way! When the company received investments, we would conduct a cashout, meaning employees could sell a small portion of their stock options for real money. Everything seemed to be going our way, like we were living in a movie. Unfortunately, though, this is the point in my story where things get a little darker, like in the Harry Potter movies. Suddenly, we were faced with a huge challenge. In 2020, the world was shut down by the coronavirus. inDrive closed their offices and we started to work from home—something we had always been firmly against. Life took on new rules and we had to quickly find our footing. We developed new processes, took our first online calls, and sent laptops home to new employees. Then, in mid-2020, I became the CTO of inDrive. It was a surprise even for myself. I was placed in charge of 50+ developers, 4 distributed teams, and the two offices in Moscow and Yakutsk. Of course, there were plenty of challenges, which I'll discuss in more detail in my next article. My joke about our office on Belorusskaya quickly became reality. At the end of 2020, we opened an office in one of the best business centers in Moscow, right on Belorusskaya Square, alongside other cool IT companies. P.S.

Building a Unicorn: The inDrive Tech Journey from Local Startup to Global Phenomenon 

ইউনিকর্ন তৈরি করা: স্থানীয় স্টার্টআপ থেকে গ্লোবাল ফেনোমেনন পর্যন্ত ইনড্রাইভ টেক জার্নি

Aufbau eines Einhorns: Die inDrive-Tech-Reise vom lokalen Startup zum globalen Phänomen

Construyendo un unicornio: el viaje tecnológico de inDrive desde una startup local hasta un fenómeno global

Construire une licorne : le parcours technologique d'inDrive, de la startup locale au phénomène mondial

यूनिकॉर्न का निर्माण: स्थानीय स्टार्टअप से वैश्विक घटना तक इनड्राइव टेक यात्रा

ユニコーンの構築: 地元のスタートアップから世界的現象への inDrive テクノロジーの旅

유니콘 만들기: 로컬 스타트업에서 글로벌 현상까지의 inDrive 기술 여정

Construindo um unicórnio: a jornada tecnológica do inDrive, de uma startup local a um fenômeno global

Создание единорога: технологическое путешествие inDrive от локального стартапа к глобальному феномену

Bir Tekboynuz Yaratmak: Yerel Startup'tan Küresel Fenomen'e inDrive Teknoloji Yolculuğu

Xây dựng kỳ lân: Hành trình công nghệ inDrive từ khởi nghiệp địa phương đến hiện tượng toàn cầu

打造独角兽：从本地初创企业到全球现象的 inDrive 技术之旅

If there was a time in history I would have loved to witness firsthand, one of them would have been to visit the World’s Fair, Futurama, and World of Tomorrow back in the 1930s and 40s. The Roaring Twenties, with all its decadence and "fuck you, economy" passion, gave way to a time of unbridled imagination and envisioning future possibilities with no constraints. We hadn’t even harnessed the atom at that point, emerging from one war and economic depression to rebound and reimagine the world around us filled with wonder. So what the f*ck happened to all that? Around the same time as the first set of World’s Fairs kicked off, H.G. Wells coined the term ‘foresight’. As it stands today, it still encompasses what we know as futurology, futurism (not the art form), future studies, and so on. It’s worth looking at the for the full history of how the profession grew from its origins because it moved through phases of imagining the future to providing statistical probabilities of several futures. Wikipedia entry “I began to see the future not as a totally impenetrable realm about which we can know absolutely nothing, but rather as an exciting frontier, offering enormous possibilities but also extraordinary dangers. We cannot possibly know everything that lies ahead, but with effort we can glimpse the possibilities of our future. This weak but incredibly valuable knowledge is critically important if we are to make wise decisions.” — Edward Cornish, founder of the World Future Society One of the biggest modern shapers of foresight and future studies was Edward Cornish—founder of the World Future Society. Back in the 60s and 70s he and the organization were pretty instrumental in helping to formalize a lot of the methods and tools practitioners used, bringing the profession to the no less. Unfortunately, the WFS today is nothing but a website and online forum that can only be described as worse than a subreddit. It holds no cachet or influence anymore and the legacy that Cornish spent decades building was bought out and handed to people who couldn’t future themselves out of a plastic bag. White House . The current WFS owners have sadly razed the Wikipedia entry to the ground to remove its 60-year history. There is some poetic irony in that the World Future Society only exists now in the Wayback Machine However, history lessons aside, something raw disappeared in the modernity of futurology. as the Wiki entry concludes and that is where the magic ends. “Futurists are practitioners of the foresight profession, which seeks to provide organizations and individuals with images of the future to help them prepare for contingencies and to maximize opportunities.” In a relentless pursuit to be accepted as a tool of the corporate and political world, futurists became just another bunch of McKinsey knock-offs, using their methods to ensure that their current employer can hit their KPIs and Quarterly results, or strategize for the next wild card event ( ). It’s now about predicting marketing, social trends, and consumer intents instead of dreaming and extrapolating what could be. which I may add, it seems nobody predicted Covid at all or even helped their corporate employer to prepare for a pandemic This has, unfortunately, become a problem in itself because as it becomes nothing more than trend spotting and extrapolation of percentages and blog posts so too has it attracted an unsavory bunch of charlatans who haven’t spent a single day on a foresight project, study, report, talk and so on. Throw a dart at the LinkedIn community and you’ll spear several hundred in one shot who claim to be Futurist but on checking their background you find they have nothing but Sales Assistants for Boots ( ), Fashion Futurist ( ), Planetary Futurist ( ) to Education Futurist ( ) listed. Retail Futurist Changing Room Assistants at Zara owned a florist ran an online marketing course on LinkedIn And they make more money than the practitioners because they’re not buried under reports and statistics but out there on podcasts and keynotes spitting out whatever listicles they read on a PowerPoint written by a at a creative agency. GPT model More on the GPT thing at the end. Futurism has become a cheap hobby for everyone who just looks up Google Word trends and extrapolates what the hottest thing for next year is going to be. Being a Futurist is like being the new Influencer or Evangelist. It’s a badge that tries to separate itself from the previous generation’s jaded job title that nobody wants anymore because the last mega-trending subject sullied it. It doesn’t take much to be a futurist now, just tweet/x something on a hashtag about your industry, add “AI” or “Spatial Computing” in the sentence, and voila — F U T U R I S T I C. ; you’re an evangelist. You make money selling advisory services explaining what this means for people and companies to take advantage of it and stand at conferences showing people charts like the above and making bold statements about why it will become the norm in 3 years. If you’re talking about a particular subject or trend and it’s already here then this is not foresight it is marketing This nonsense spawned exponentially during the pandemic because, suddenly, the future of work was an easy target. People and execs panicked, and they didn’t know what the fuck was going to happen, so why not ask someone who calls themselves a Futurist who talks about the Future of Work, virtual work meetings, and metaverse workplaces? They’ll know how to handle this. And then you have various strains of futurists, all trying to differentiate themselves by inventing something else to appear hip. For context: — someone who can’t tell you about future possibilities but can talk all day long about what’s happening today. Nowist — someone who can’t tell you about future possibilities but can talk all day long about what’s happening tomorrow. Near Futurist — someone who failed to get into Accenture but can talk all day long about strategy. Practical Futurist Shit, even I went by the title at one point to get away from it all. Now I feel like a , a term derived from , a word I found in a book that explained the feeling and lamentation of promised and imagined futures that failed to materialize. Anti-futurist Hauntologist Hauntology So, now even the practicing futurists are ever so subtly removing or tweaking the title from their public profiles because there is little to be gained from the association with the term. And you can’t criticize the profession. That’s bad juju. They’ll hunt you down with pitchforks behind the membership paywalls. You can’t talk about the bad stuff and negative consequences in futurist outlooks openly either because futurism is supposed to be about tripping the light fantastic to utopia whilst you suck on some techno-optimists teat. I was asked several years ago to give an end keynote at an event in Peru on artificial intelligence. Make it happy they said. I didn’t, the realities of extrapolating many different scenarios means you can’t just tell them the fluffy stuff because that doesn’t prepare them for the economic reality — especially in a region that isn’t a typical Westernised first-world clone of Silicon Valley. You are bound to explain everything. This is what separates some futurists from each other. There are those who will only spit the current narrative or positive Valley rhetoric because it will secure more work and make them more appealing to engage with as an influencer for the corporate shilling, and there are those who would rather tell it like it could be, warts and all with no compromise. The very sad part though, is that there is no room for real imagination anymore in this brave new world of future studies. If you take a look at the various narratives across futurology and possibilities of the future they are nowhere near as grandiose or as visionary as one hundred years ago. but now visions are cocooned and constrained by technology and trend spotting and their perception of progress and change. There is nothing from the late 19th and early 20th century that wasn’t imagined for today, just the timelines were always a bit optimistic and funky. You can even find old cartoons of people using video-calling devices drawn in late 1800s. Back then, imagination was unbounded and unconstrained, unfettered by numbers and correlation But now? No-one is imagining, only copying what they’ve read in a Gartner report. Far too many people think that imagination is tied to money or investment or needs an existing technology to back it up. This is the fucking problem now. “I can’t imagine something big or wild because people won’t believe it or invest in it.” Who fucking cares?! We’ve lost the ability to dream better futures because we’ve been conditioned to believe there has to be a pitch deck that accompanies it. Ask a TikTok futurist and you’ll get the usual crap like “we’ll all be wearing augmented reality glasses in the next 5 years” or “AI in healthcare and robotic surgery” or “digital culture in the metaverse”. Even the old guard like Peter Diamandis and Ray Kurzweil have given up trying and regurgitating the same, tired ideas about nano-this and quantum-that. Is there nothing new to imagine in futurism? There is a little glimmer of hope though. There is a method called that combines the past with the present, including interviews with notable science fiction authors to provide the tools needed to “design the future with science fiction” and this is where the real future and direction of futurism lies, in a sense going back to its roots in science fiction and H.G. Wells to imagine bigger possibilities, both good and bad and putting them on display to inspire the public like we did before. Science Fiction Prototyping And no, the abomination that is CES does not count ffs. To combine outrageous imagination with the various methodologies that sit within future studies could offer, oddly enough, potential futures in itself for the profession that is so stuck in the past that it has no real future. , I interviewed a primary school year for a project and what they came up with was more fun, probable, positive and imaginative than those sitting on $500k contracts. Even a school child imagines bigger and brighter visions of the possible and the probable and I should know . But we have lost the drive to put our thoughts out there because the current trend is to shoot anything down that doesn’t remotely feel real or achievable or backed by something you read on Arxiv. And so we end up in this cycle of repeating the same crap over and over but using the latest marketing term to keep the prediction fresh (see the trend above chart above as a reminder) Being a futurist should be more about Gedankenexperimente than statistics So now we near the end of this rant. Didn’t see it coming, did you…not much of a futurist are you, but then again you’re exactly what futurism does need — someone who hasn’t been soiled by the stuffy academia of it all or who doesn’t care about the keynote glamour but is genuinely interested in the agency they possess to not only imagine and inform but shape the future they want to see at the same time. We will have positive and utopian and harmonic visions, protopian visions of an inclusive future, and dystopian visions that lead us to hell. The trick is knowing how to plan and prepare for every one of them and their possibilities and dangers, and to not be afraid to speak about them however harsh. For every dystopian future, there is a positive one to imagine out of it that could occur from a single decision we all possess the power to influence but you must know hell to recognise heaven, and that’s the real power of futurism. So, where have all the futurists gone? We’re still out there. Waiting. You might catch a glimpse of one in the corner of your eye at a conference, but if you turn to look they’ve vanished. Or try to take a photo of one but all that comes back is a blurred photo despite owning a 108MP smartphone camera. I’ll end this piece by stating two things. (1) There are not many futurists I respect anymore but those I do are all women. (2) I scared some people by saying I trained a ChatGPT model for future studies and it was pretty effective once you knew how to use it. And that in itself tells you why we need to go back to the future. Also published here

Will The Real Futurists Please Stand Up?

197 Stories To Learn About Startup Marketing

EASE NOT PEACE.

In this article, we will explore the concept of clustering in machine learning, delving into the different types of clustering and examining various algorithms within the clustering technique. Before delving into the specifics of clustering, let's take a moment to survey the different types of learning in : machine learning Supervised Learning: In supervised learning, models are trained on labeled data, where the term "label" corresponds to the category of the input data. For instance, consider a spam classification system. Here, machine learning models undergo training on both spam and non-spam emails, each tagged with their respective label tags. Tasks such as classification and regression fall within the purview of supervised learning. Unsupervised Learning: Conversely, unsupervised learning stands in contrast to supervised learning, operating without any labeled data. In this approach, models are trained on data devoid of explicit labels. Take, for example, the realm of recommendation systems, which operates on unsupervised learning techniques. Clustering, as a task, falls under the umbrella of unsupervised learning. While other learning types, such as reinforcement learning and semi-supervised learning, exist, they are not the focus of our discussion in this article. Building upon the above explanation, it becomes evident that clustering falls under the domain of unsupervised learning, where the data provided to the model during training is unlabelled. Now that we have a contextual understanding of where the concept of clustering emerges, it's time to delve into the essence of what clustering truly is. What is Clustering? In simple terms, clustering is a technique that involves grouping data based on certain characteristics. These groups of data points are referred to as "clusters." To illustrate this concept, let's explore a small example. Consider the following sentences as a dataset: Cricket is a fun game. Cars are cool machines that help us travel from one place to another. Teams take turns to bat and bowl. Scoring lots of runs makes you a hero in cricket! They have wheels, an engine, and can go really fast on roads. Some cars are big like trucks, and others are small like tiny race cars. By examining the properties and meanings of these sentences, we can group them into two categories: cricket and cars. This grouping aligns with what a clustering algorithm accomplishes during its training process. It captures the context (meaning) of the input data in numerical (vector) form and clusters the data points (vectors) based on their similarity. During inference, the clustering algorithm calculates the similarity between a query vector and the formed clusters (established during training) and assigns the query vector to the closest cluster. We will delve into the specifics of similarity metrics later in this article. How does the Clustering algorithm work? While a brief introduction on how clustering works was provided earlier, let's now delve into the procedural perspective. In the realm of machine learning, various clustering algorithms exist. However, the following procedure is a universal framework applicable to any clustering algorithm in use. Data Preprocessing: In the initial phase, text data undergoes preprocessing techniques like tokenization, lemmatization, and stop-word removal. This step enhances the quality of the data for subsequent analysis. Embedding Generation: Following preprocessing, we proceed to generate embeddings (vectors) for the processed text data. Various approaches exist for this task, ranging from traditional techniques like bag-of-words to more recent advancements such as transformers and language model pre-training. Cluster Formation: During the training of the clustering algorithm, clusters are formed based on the similarity between vectors. Metrics like cosine similarity or Euclidean distance are employed to calculate the similarity between vectors. It is essential for a robust clustering algorithm that intra-cluster distance is minimized while inter-cluster distance is maximized. Assignment of New Data Points: In the inference phase, the clustering algorithm compares a query vector to each cluster and assigns it to the cluster where the similarity score is high or the distance is low. This step ensures that new data points are accurately placed within existing clusters based on their similarity to cluster centroids. Types of Clustering There are different approaches to clustering a group of data points based on the similarity in their vector representation. A few of the techniques are discussed here : Hierarchical clustering Centroids-based Clustering Density-based Clustering Hierarchical clustering Hierarchical clustering is also known as Connectivity-based Clustering, where data points are connected to neighbors based on their relationships (proximity distance) The clusters are represented in a graph-like structure called a Dendrogram. The X-axis of the dendrogram represents the objects that do not merge, while the Y-axis represents the distance at which clusters merge. Similar data objects have minimal distance, falling into the same cluster, while dissimilar data objects are placed farther in the hierarchy. There are two most common types of Hierarchical Clustering: Agglomerative Clustering: Also known as the AGNES algorithm, this type of clustering adopts a bottom-up approach. Initially, all data points are treated as individual clusters, resulting in N clusters. Through an iterative procedure, these N clusters gradually merge with neighbouring clusters based on their similarity. By the end of the iterative process, all the individual clusters are amalgamated into a single large cluster containing N data points. Divisive Clustering: Also known as the DIANA algorithm, divisive clustering takes a top-down approach. All data points start within a single large cluster. Through an iterative process, this massive cluster breaks into sub-clusters and continues until a termination condition is met. This approach contrasts with agglomerative clustering, providing a different perspective on hierarchical clustering methodologies. Centroids-based Clustering This clustering technique, also recognized as Partition-based Clustering, stands out as one of the simplest and most widely utilized methods in the realm of unsupervised learning. The renowned k-means clustering algorithm is a prominent example within this category. In this approach, datasets are partitioned into a predetermined number of clusters, with each vector of values associated with a specific cluster. The process involves comparing a new vector with all cluster centroids and assigning it to the cluster with the minimum distance. This method offers a straightforward yet effective means of organizing data into distinct clusters. Common distance metrics used for calculating cluster distances in this type of clustering technique include Euclidean distance, Manhattan distance, or Minkowski distance. However, a notable limitation of this clustering approach is that the initial number of clusters (k) must be defined beforehand. This requirement can pose a challenge, as determining the optimal value for k may not always be straightforward and may require additional considerations or techniques. Density-based Clustering In contrast to the aforementioned clustering techniques, density-based clustering does not rely on distance metrics but instead considers the density of the data points' distribution. Data points are clustered based on their densities, wherein areas with a high concentration of points are identified as dense clusters, while low-concentration areas are considered separate clusters. Another notable distinction from the previously discussed techniques is the effective handling of outliers by density-based clustering algorithms. Outliers are designated as such, providing a more flexible approach compared to centroid-based clustering, where outliers are compelled to be assigned to one of the predefined clusters. Prominent algorithms in this clustering technique include DBSCAN and OPTICS. These methods showcase the adaptability of density-based clustering in capturing intricate patterns within datasets. Applications of Clustering Let's explore scenarios where clustering becomes a preferred choice over other techniques in machine learning. Below are specific use cases where clustering proves to be highly effective: Unlabelled Data or High Labelling Costs: When the available data is unlabelled, or the cost associated with labeling the data is prohibitively high. Scenario: Clustering is particularly valuable in scenarios where obtaining labeled data is impractical or costly. Significance: Recommendation and Search Engines: Tasks such as recommendation engines and search engines leverage clustering techniques. Scenario: Netflix employs clustering for movie suggestions, demonstrating the practical application of clustering in enhancing user experience and content discovery. Example: Customer Segmentation: In cases where businesses need to segment customers or markets based on the sales of a particular product or service. Scenario: Clustering aids in identifying distinct customer groups, allowing businesses to tailor their strategies based on specific customer characteristics. Significance: Anomaly Detection in Finance and Banking: Detecting outliers and anomalies in financial data within the banking sector. Scenario: Clustering serves as a robust tool for identifying unusual patterns or transactions that deviate from the norm, contributing to enhanced fraud detection and risk management. Significance: Document Classification and Retrieval: Classifying and retrieving documents based on their content. Scenario: Clustering techniques are instrumental in organizing and categorizing large volumes of documents, streamlining the document retrieval process. Significance: Real-Time Problem Solving: Clustering is applied to address various real-time problems. Scenario: The versatility of clustering techniques makes them adaptable to a range of dynamic, real-world challenges, providing effective solutions in diverse contexts. Significance: Conclusion In conclusion, clustering emerges as a versatile tool, particularly in scenarios involving unlabelled data or where labeling proves cost-prohibitive. Its broad utility extends across various domains, including recommendation systems, customer segmentation, anomaly detection, document classification, and real-time problem-solving. For , clustering stands as a valuable asset within their toolkit, offering a powerful means to extract meaningful insights from complex and unlabeled datasets. machine learning practitioners

Introduction to Clustering in Machine Learning: Types, Algorithms, and Applications

India's obsession with and digital marketing scams is indeed exacerbated by the high number of people pretending to be SEO professionals and experts, while in reality, they are nothing more than individuals engaged in SEO reseller programs. SEO These self-proclaimed "experts" are often far from it, and they contribute to the erosion of trust and professionalism in the industry. These so-called SEO professionals often lack the in-depth knowledge and skills required to deliver effective SEO services. They claim expertise in search engine optimization, but, in truth, they are merely middlemen, outsourcing the actual work to others while pocketing a significant margin. They rely on deceptive marketing tactics to portray themselves as experts, taking advantage of the widespread fascination with SEO in India. Who’s to blame? The blame for this situation extends to the agencies that offer these reseller programs. By relying on individuals who are willing to deceive clients about their true capabilities, these agencies become the core of the problem. They often prioritize their own profits over the quality of service they provide, creating a breeding ground for scams and unethical practices. This business model not only perpetuates the deception but also devalues the true expertise that legitimate SEO professionals possess. It's crucial to start to expose and condemn this unethical behavior on both sides—those masquerading as experts and the agencies enabling them. The lack of transparency and accountability in these reseller programs is a major issue, and it is high time the industry and regulatory bodies take action to address this problem. It’s of my own personal opinion the presence of individuals pretending to be SEO experts in India, often engaged in reseller programs, exacerbates the country's obsession with SEO and digital marketing scams. To combat this issue, we must not only scrutinize these pretenders but also hold the agencies offering reseller programs accountable for their part in perpetuating this deceptive culture. The Rise Of Digital Marketing Scams In India You won't believe the audacity of the situation in 2022 - there has been an astronomical 80% rise in the number of fake digital marketing "experts" in India! It's like a plague that has infected the digital landscape, and it's high time we called them out on their dubious practices. Everywhere you turn these days, you're bombarded with self-proclaimed digital marketing gurus who promise to revolutionize your online presence. They dangle the carrot of increased website traffic, more leads, and a sales bonanza right in front of you. But let's not kid ourselves; most of them are nothing but charlatans preying on the unsuspecting. These so-called "experts" are like the snake oil salesmen of the digital age. They talk a big game, but when it comes down to delivering real results, they're about as effective as a screen door on a submarine. It's infuriating how they manipulate their way into the industry, peddling their pseudo-knowledge and empty promises. The rise of digital marketing in India has given birth to a monstrosity that's plaguing businesses everywhere - the surge of fake digital marketing "experts." These self-proclaimed gurus have invaded the market, and it is high time we addressed this menace. Picture this: you're in desperate need of digital marketing expertise. You want to boost your online presence, drive traffic, and generate leads. So, you start hunting for a digital marketing pro. And what do you find? A sea of charlatans who talk the talk but can't walk the walk. They claim to possess digital marketing wizardry but lack the skills and knowledge to back up their lofty assertions. What's fueling this digital deception, you ask? It's the harsh reality of supply and demand. The scarcity of skilled digital resources in India has led to a grotesque imbalance! With genuine experts being as rare as a unicorn, fake digital marketers have seized the opportunity to charge exorbitant prices for their dubious services or provide little to no value. But wait, there's more to this digital marketing nightmare! The problem is compounded by the complete absence of a regulatory body overseeing digital marketing in India. In this Wild West of the online world, virtually anyone can hang their shingle and declare themselves a digital marketing expert. Yes, you read that right – there are no rules, no regulations, and no one to hold these self-proclaimed gurus accountable for their actions. It's the digital marketing equivalent of the lawless frontier. This glaring gap in oversight has allowed these charlatans to run amok, preying on unsuspecting businesses without consequence. It's a free-for-all, where anyone with a computer and a persuasive pitch can don the mantle of a digital marketing "expert." There are no qualifications, no standards, and no quality control. As a result, businesses are left vulnerable, and they're often at the mercy of these self-appointed swindlers. The absence of a regulatory body for digital marketing in India is a glaring loophole that needs to be addressed urgently. We can't continue to let these fake digital marketing people thrive in this lawless environment. I personally demand accountability, standards, and genuine expertise in the world of digital marketing. The Flock To Freelance Platforms - Scam Home Base I absolutely cannot believe the audacity of some of these people flocking to freelancer services like Fiverr and Upwork to conduct digital marketing scams! It's an absolute disgrace, and it tarnishes the reputation of these platforms, as well as the entire digital marketing industry. In fact, I noticed the scams were getting so bad I actually took it upon myself to learn as much of the science of SEO as possible and do everything myself. It’s funny. I actually learned SEO from the scammers themselves (What not do to anyway). Let's talk about the scammers themselves. These charlatans are preying on unsuspecting individuals and businesses who are genuinely looking for help with their digital marketing needs. They promise the moon and the stars, , social media, content marketing, and more. They lure their victims in with cheap prices and promises of quick results. But in reality, they have no clue what they're doing, and their tactics often violate the rules and guidelines of reputable platforms. claiming to be experts in SEO And what's even worse is that many of these scammers get away with it! They use fake credentials, fabricated reviews, and stolen portfolio samples to create an illusion of expertise. People fall for their deceit, only to find out later that their hard-earned money has been wasted on empty promises and ineffective marketing strategies. This not only harms the individuals who've been scammed but also undermines the trust that should exist in these freelancer platforms. But it's not just the scammers who are at fault. It's also the responsibility of these platforms to step up their game and protect their users. They need to implement better vetting processes, stricter review policies, and more comprehensive background checks to filter out the bad actors. The fact that these scammers can continue to operate on these platforms is unacceptable. In the end, this wave of on platforms like Fiverr and Upwork is a stain on the industry and a disservice to honest freelancers and clients alike. digital marketing scams It's time for everyone involved to take responsibility, from the scammers and the platforms to the clients who enable this behavior. The digital marketing world deserves better than this. Genuine professionals and ethical practices in the SEO and digital marketing industry should be championed and celebrated to restore trust and credibility in the field.

India's Obsession With SEO And Digital Marketing Scams 

Is a Career in Game Development Right for You? 

গেম ডেভেলপমেন্টে ক্যারিয়ার কি আপনার জন্য সঠিক?

Ist eine Karriere in der Spieleentwicklung das Richtige für Sie?

¿Es adecuada para usted una carrera en desarrollo de juegos?

Une carrière dans le développement de jeux est-elle faite pour vous ?

क्या गेम डेवलपमेंट में करियर आपके लिए सही है?

ゲーム開発のキャリアはあなたに適していますか?

게임 개발 경력이 당신에게 적합합니까?

Uma carreira em desenvolvimento de jogos é ideal para você?

Подходит ли вам карьера в разработке игр?

Oyun Geliştirmede Kariyer Size Uygun mu?

Nghề phát triển trò chơi có phù hợp với bạn không?

游戏开发职业适合您吗？

Business English: A Practice Book by Rose Buhlig is part of the HackerNoon Books Series. You can jump to any chapter in this book . THE PREPOSITION AND THE CONJUNCTION here THE PREPOSITION AND THE CONJUNCTION **Prepositions
\ It is important in the study of prepositions to observe that there are certain words that are followed by certain prepositions. To change the preposition is to convey a different meaning from the one that the speaker intended, or to convey no meaning at all. A partial list of such words with their appropriate prepositions follows: accompanied with anything having no life accompanied by anything having life acquit of accuse of adapted to a thing adapted for a course, because of one's nature adapted from an author agree to a plan or proposition agree with a person agree upon something that must be decided angry at a thing angry with a person compare with to bring out similar qualities compare to without analyzing comply with confer on meaning to give to confer with meaning to talk to confide in meaning to put faith in confide to meaning to commit to one's keeping conform to correspond to, with a thing, denoting similarity correspond with meaning to write to different from dependent on a person dependent for a thing independent of disappointed in employed at a certain place or salary employed in a certain kind of business employed by a certain person or company enter upon duties enter at a door followed by influence over, upon expect of participate in profit by remonstrate against a thing remonstrate with a person Insert the correct preposition in the following: Exercise 125 I shall comply —— your request. The chairman came upon the platform accompanied —— the speaker. He took a walk accompanied —— his dog. The lecture will be accompanied —— stereopticon views. Strikes are usually accompanied —— riots. The years of prosperity were followed —— years of famine. He was accused —— theft, but was acquitted —— the accusation. She is well adapted —— the position that is open. An electric iron is especially adapted —— summer use. The selection was adapted —— Irving. This cloth is well adapted —— summer clothing because it is very light in weight. I agree —— you that the plan is impracticable. Let us agree now —— a place to spend our summer vacation. That is not a proposition —— which I shall agree. It is silly to be angry —— an inanimate object. Don't be angry —— a person because he tells you your faults. His report corresponds in all respects —- yours. Mr. Giles suggested that you would be glad to have us correspond —— you concerning our new bond issues. I shall confer —— my lawyer. The public has conferred a great honor —— him. One should always profit —— his experiences. The new device is entirely different —— the old. I am employed —— a fairly large salary —— a business that is growing daily. All employees must conform —— the rules. I am confiding —— you because I know that I can trust you. She confided her child —— the care of her brother. She is dependent —— her brother —— support. You can have an influence for good —— him. I have remonstrated —— the change several times. Perhaps he will change his plans if we remonstrate —— him at once. Exercise 126—Prepositions Incorrectly Used Each of the incorrect sentences given below contains an unnecessary preposition. When the meaning of "Where are you going?" is entirely clear, there is nothing gained by saying "Where are you going " Omit such superfluous prepositions. to? Wrong Right 1. I took it off the shelf. of I took it off the shelf. 2. I shall accept your hospitality. of I shall accept your hospitality. 3. Where are you at? Where are you? 4. Where are you going to? Where are you going? 5. It is a building twenty to thirty stories in height. of from It is a building twenty to thirty stories in height. 6. Look out the window. of Look out the window. 7. John copies his father in everything. after John copies his father in everything. 8. I am wondering what I should do. about I am wondering what I should do. 9. I shall consult my lawyer. with I shall consult my lawyer. 10. He sat opposite me. to He sat opposite me. 11. I shall leave later . on I shall leave later. for and to 12. I shall try go. and I shall try go. to for of have 13. I might gone. of I might gone. have The wrong preposition 14. He fell the water. in He fell the water. into 15. She died diphtheria. with She died diphtheria. of 16. Divide the work the four of us. between Divide the work the four of us. among ( may be used in speaking of only two persons or things) Between 17. It will be done of an hour. inside It will be done an hour. within 18. Are you angry me? at Are you angry me? with Preposition must be used 19. It's no use to try. It's no use to try. of 20. My sister stayed home. My sister stayed home. at 21. Why do you act that way? Why do you act that way? in 22. We left the third of June. We left the third of June. on The object of a preposition is always in the objective case. Some people have great difficulty in recognizing that in such expressions as for you and me, the pronoun me is as much the object of the preposition for as the pronoun you. Both words must be in the objective case. It is incorrect to say for you and I. Exercise 127 In the following sentences omit the incorrect italicized form: The invitation is for father and I—me. Every one has finished his work except he—him and I—me. It's a question that you and I—me must decide; it refers to you and I—me alone. Girls like you and she—her should have a good influence over the others. All but you and I—me have left. He did it for you and I—me. No one objected but they—them and we—us. She sat opposite you and I—me. They were sitting near you and I—me. We expect you to return with mother and I—me. He wanted my brother and I—me to go into business with his brother and he—him. Neither she—her nor her sister have I seen for several months. My companion and I—me took up the trail of the bear at once. For some distance it led he—him and I—me over the soft, yielding carpet of moss and pine needles, and the footprints were quite easily made out. He—him and I—me had, of course, to keep a sharp lookout ahead and around for the grizzly. All are going on the excursion except he—him and I—me. He—him and I—me went fishing. The rule applies to we—us all—the manager, they—them who keep books, you, and I—me. She beckoned to my companion and I—me. The letter was to be read by the president or I—me. He did it for the sake of my father and I—me. We study Shakespeare with her sister and she—her. She—her and her sister went to the lecture with my sister and I—me. They sent for she—her and I—me, not you and he—him. The program was arranged by the president and I—me. They found that his father and he—him had already left. Mother is going to buy a birthday present to-day for she—her and I—me. The play is interesting not only to you older people but to we—us younger ones also. They expected the work to be done by she—her and I—me. The dispute between his neighbor and he—him over their lot line was settled by the surveyors this morning. He wants to speak to you and I—me. Than and as are not prepositions but conjunctions. They are used to introduce subordinate clauses. Usually the clause is incomplete, but the omitted part is easily understood from the preceding clause and must be supplied to show the case of the noun or the pronoun that is expressed; as, Exercise 128—Than, as : She is as tall as I am. Right : She is taller than he is. Right : I should invite you rather than her than I should invite her. Right Use the correct one of the italicized pronouns in the following sentences: I'll agree that he is richer than I—me, but riches are not everything. I shall send her rather than he—him. No one felt sorrier than she—her. No one knows more about an automobile than he—him. You are more capable of doing the work than he—him. We were nearer the goal than you or he—him. You finished the work almost as quickly as she—her. She writes fully as well as he—him. The manager said he would rather send me than he—him. I secured a position sooner than she—her. It seems to me that they ought to go rather than we—us. I am surprised that you arrived sooner than they—them. They should have elected him rather than I—me. I am not so well-fitted as he—him to hold the position. You are more popular than he—him. Exercise 129—Correlatives There are certain conjunctions, called , that are used in pairs. They are correlatives both—and as—as, so—as either—or not only—but also neither—nor whether—or so—that such—as Illustrations Both—and He has both skill and energy. Either—or I shall leave either Monday or Tuesday. Neither—nor I can neither sing nor play. So—that It rained so hard that we stayed at home. As—as We shall come as early as we can. So—as She is not so tall as you are. (Used in negative expressions.) Not only—but also We saw not only Mr. Brown but his wife also. Whether—or Whether I return to work or stay at home depends on my mother's health. Such—as We shall buy only such goods as we think we can sell. Be very careful not to use the correlative so as incorrectly for so that. So as is used in negative expressions of comparison; so that is used to express result. : We went early so as we could get good seats. Wrong : We went early so that we could get good seats. Right In the illustrations given above, notice that the correlatives always join two similar or coördinate expressions. It is important that they be placed each immediately before one of the two coördinate expressions. : I neither can sing nor play. Wrong : I can neither sing nor play. Right Recast the following sentences, placing the correlative conjunctions before coordinate expressions: Either you ordered it late or not at all. He said he neither had money nor time. We not only bought the books you wished but the games also. We like the place in which we live both on account of its quietness and its pleasant surroundings. I shall either go to Quebec or Montreal. Either he must spray his trees or expect no fruit. I neither like the appearance of the shop nor the attitude of the clerks. They did it both for the sake of your brother and you. This sample not only is much darker but heavier also. They are barred who neither can read nor write. These conjunctions are correctly used in speaking of two things only. Care must be taken to use or with either and nor with neither. In comparing three or more things use any of them, none of them, or no. Exercise 130—Either—or, Neither—nor In the following sentences use only the correct italicized forms: Neither effort nor—or money was spared in the undertaking. I have considered planting maple, oak, and elm trees, but neither—none of them seems to grow well in this climate. We do not believe in either enduring oppression nor—or killing the oppressor. We believe in arbitration. He has no—neither time, patience, nor—or energy. If you ask me which of the three I prefer, I'll be frank and tell you I like neither—none of them. Three courses will be given in the subject this year; you may take either—any one of them. I had already passed three branch roads, but neither—none of them had looked familiar to me. I hardly think he accepted any—either of the two offers he received. Neither the doctor or—nor his wife was at home. Both the books look shop-worn. I'll take neither—none. Except and without are prepositions, and are used, therefore, to introduce phrases; unless is a conjunction, and is used to introduce a clause. Exercise 131—Except, Without, Unless In the following sentences insert the correct form, giving a reason for your choice: —— you leave at once, you will miss your train. I cannot learn to swim, —— some one teaches me. I cannot learn to swim —— a teacher. No one could do the work —— me. John expects to learn —— studying. John will discover that he cannot win promotion —— he works hard. No one can learn how to spell —— first learning how to observe. No one will learn to spell —— he learns to observe. No one will succeed —— he has energy and patience. No one will succeed —— energy and patience. You cannot succeed in any way —— by seizing each opportunity as it comes. It is impossible to grow beautiful flowers —— the soil is good. Exercise 132—Like, as Like is followed by a noun or pronoun in the objective case. As is a conjunction and introduces a clause, and is therefore followed by a verb. Like is not a conjunction and therefore may not be substituted for as or as if. : I wish I could play like you can. Wrong : I wish I could play as you can. Right Insert the correct word in the following sentences: The picture looks just —— you. I haven't a voice —— my brother's. I cannot sing —— my brother can. He walks just —— you do. I hope you will all enjoy the trip —— I did. For pleasure and exercise I think there is no game —— tennis. He said that the town looked just —— it had when he was a boy. I cut the paper just —— you said I should. He talks —— his father. He has the same sort of drawl —— his father has. She was there —— you said she would be. They worked —— beavers. He looked —— a tramp. To give the stitch the proper twist throw the thread over the needle —— I do. He walks —— he were lame. Exercise 133—As—as, So—as Use as—as in stating equality; use so—as in negative comparisons. You will find the new clerks fully —— courteous as were the old. You will not find the new clerks —— courteous as were the old. Elms do not grow —— well in this climate as do poplars. We did not carry —— much advertising this year as we did last year, and we find that our receipts are smaller. Under our system of individual instruction a student may advance —— rapidly as his ability permits. You are not —— tall as your sister. I do not seem to learn languages —— easily as mathematics. This house is not —— large as the other. He is —— active as he was twenty years ago. He is not —— active as he was twenty years ago. Exercise 134—Miscellaneous Blunders To, Too, Two To is a preposition; too is an adverb, and means excessively or also; two is a numeral adjective. Insert the correct form in each of the following sentences: The —— sisters discovered that it was —— late for the 4:15 train. It is —— dark in that corner; come —— the light. He spends —— much time in dreaming, —— little in working. He would have done better if he had not given —— little heed —— the advice of his —— older brothers. —— more hours were passed in the all —— weary task of waiting. It was —— cold —— stay out more than —— hours. You may go ——, but don't stay —— long. —— stay there for —— weeks would be —— tiresome. The doctor said that the —— men were —— sick —— go home alone, and I thought so ——. About —— hours ago I met Mary who said that she was going —— the country ——. There, Their 11. —— are seven brothers in —— family. —— books are —— on the table. —— is no doubt that —— knowledge of mathematics is greater than —— knowledge of English. Were, Where 14. —— have you been? —— you ever on a farm —— alfalfa is grown? —— —— you when the report was read? I was just —— you ——. Of, Have 18. You should —— read more distinctly. I could —— done the work if I had had more —— the necessary tools. If I had tried harder, I might —— done the work better. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Rose Buhlig (2011). Business English: A Practice Book. Urbana, Illinois: Project Gutenberg. Retrieved https://www.gutenberg.org/cache/epub/38046/pg38046-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at , located at . www.gutenberg.org https://www.gutenberg.org/policy/license.html

THE PREPOSITION AND THE CONJUNCTION

ALLWHERE

allwhere.co

11-50 employees

Since 2021

COMPANY RANKING

ALLWHERE

EVERGREEN INDEX #1563

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

allwhere WIKI

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

ALLWHERE

allwhere.co

11-50 employees

Since 2021

COMPANY RANKING

ALLWHERE

EVERGREEN INDEX #1563

RELATED COMPANIESRANK

STORIES ABOUT ALLWHERE

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

allwhere WIKI

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