<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Kannan Nagasamy on Medium]]></title>
        <description><![CDATA[Stories by Kannan Nagasamy on Medium]]></description>
        <link>https://medium.com/@kannankannan18?source=rss-c6778e363b04------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*zj_XRBZru3kUrBxH0D_HEg.jpeg</url>
            <title>Stories by Kannan Nagasamy on Medium</title>
            <link>https://medium.com/@kannankannan18?source=rss-c6778e363b04------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 18:52:18 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@kannankannan18/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[ OpenAI Deployment Company — The $10 Billion AI Consulting Firm That Just Scared Every IT Giant…]]></title>
            <link>https://medium.com/@kannankannan18/openai-deployment-company-the-10-billion-ai-consulting-firm-that-just-scared-every-it-giant-9930567d111a?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/9930567d111a</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Tue, 12 May 2026 11:51:23 GMT</pubDate>
            <atom:updated>2026-05-12T17:17:29.299Z</atom:updated>
            <content:encoded><![CDATA[<h3>🤖 OpenAI Deployment Company — The $10 Billion AI Consulting Firm That Just Scared Every IT Giant on the Planet</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DvuoaC8ozx9Z3h2u1_9PGA.png" /><figcaption>OpenAI Deployment Company vs IT Services Companies</figcaption></figure><blockquote>What OpenAI’s new Deployment Company means for TCS, Infosys, Wipro — and every enterprise client in between.</blockquote><p>For decades, if a Fortune 500 company wanted to modernize its operations with technology, the playbook was simple: hire TCS, Infosys, or Accenture. Ship the project to a team of engineers. Wait 12–18 months. Pay millions. Repeat.</p><p>On May 11, 2026, OpenAI tore up that playbook. 🔥</p><h3>🏢 What Exactly Is the OpenAI Deployment Company?</h3><p>The <strong>OpenAI Deployment Company</strong> is a new $10 billion joint venture — majority-owned by OpenAI — designed to do something deceptively simple: go inside your company and <em>build AI for you</em>.</p><p>At the heart of the model are <strong>Forward Deployed Engineers (FDEs)</strong> — a specialized team of AI engineers who embed <em>on-site at client organizations</em>, work shoulder-to-shoulder with your business leaders, and build production-grade AI systems connected to your own data, tools, and workflows.</p><p>Think of it as OpenAI’s answer to McKinsey — but instead of PowerPoint decks, they leave behind working AI systems. 🎯</p><p>The venture was launched with backing from <strong>19 investment partners</strong>, including TPG, Bain Capital, Brookfield, Goldman Sachs, SoftBank, and Warburg Pincus. To jumpstart the team, OpenAI also acquired <strong>Tomoro</strong> — an applied AI consulting firm with ~150 engineers who’ve already deployed AI at companies like Tesco, Virgin Atlantic, and Supercell.</p><p>OpenAI’s COO <strong>Brad Lightcap</strong> leads the effort, reporting directly to Sam Altman. The message is clear: this isn’t a side project. It’s a core bet on OpenAI’s future.</p><h3>⚠️ Why Every IT Service Company Should Be Worried</h3><p>When the announcement dropped, the market didn’t wait around.</p><p>The <strong>Nifty IT index fell 3.6%</strong> in a single day — hitting its lowest point since May 2023. TCS dropped ~4%. Persistent Systems fell nearly 5%. Infosys, HCL, and Wipro were all down 2.5–4%. By May 2026, the Nifty IT index was down nearly <strong>25% for the year</strong>. 📉</p><p>Here’s why investors panicked:</p><p><strong>🎯 OpenAI is moving from “pick and shovel” provider to full-service contractor.</strong> IT firms have built a $250B+ industry on helping enterprises design, build, and maintain technology systems. OpenAI’s Deployment Company is now doing exactly that — just faster, cheaper, and with frontier AI baked in.</p><p><strong>🤝 The PE investor network is a distribution machine.</strong> The 19 PE firms backing OpenAI collectively have portfolio companies spanning <strong>2,000+ businesses</strong>. That’s a ready-made client pipeline that would take a traditional IT firm decades to build. OpenAI skips the 18-month sales cycle entirely.</p><p><strong>⚡ Speed is the killer feature.</strong> While a traditional IT engagement takes 6–18 months to deploy, FDEs can have production AI systems running in <strong>4–8 weeks</strong>. For a CFO counting quarters, that’s not a small difference.</p><p><strong>💰 And then there’s the pricing model…</strong></p><h3>💸 The Numbers Don’t Lie: Service Firm vs. OpenAI Deployment Co.</h3><p>Let’s use a real scenario. <strong>Company ABC</strong> wants to automate customer support with AI, handling <strong>5 million interactions per year</strong>.</p><h3>📊 Year-by-Year Cost Breakdown (Rough analysis)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/552/1*ER2zI21u2AD170P4ksL_yg.png" /></figure><p>🧮 <strong>The API cost math:</strong> 5M interactions × 2,000 tokens = 10 billion tokens. At GPT-5.4 pricing ($2.50 input + $15 output per 1M tokens) = <strong>~$175,000/year</strong>.</p><p>💡 <strong>The outcome-based fee:</strong> AI saves $500K/year in support costs. OpenAI takes 20% of that = <strong>$100K/year</strong>. You keep the other $400K.</p><h3>🏆 The Verdict</h3><p>OpenAI’s model is <strong>~26% cheaper over 3 years</strong> — saving Company ABC roughly $735,000. The Year 1 gap is enormous ($745K difference) because OpenAI eliminates the expensive bespoke development phase entirely.</p><p>But here’s the catch: by Year 2, costs nearly converge. And at very high volumes (50M+ interactions/year), OpenAI’s API bill alone could exceed $1.75M/year — making the traditional fixed-price contract look attractive again. 🤔</p><h3>✅❌ The Honest Pros and Cons</h3><h4>👍 Why You’d Choose OpenAI Deployment Company</h4><ul><li><strong>⚡ Speed:</strong> Live AI systems in weeks, not months</li><li><strong>💰 Lower upfront cost:</strong> No massive Year 1 development spend</li><li><strong>🔄 Always-current models:</strong> Automatic access to the latest frontier AI — no re-contracting needed</li><li><strong>🤝 Shared risk:</strong> Outcome-based pricing means OpenAI only wins when you win</li><li><strong>🧠 Frontier intelligence:</strong> Your systems are powered by the world’s most capable AI models</li></ul><h4>👎 Why You Might Think Twice</h4><ul><li><strong>🔒 Vendor lock-in:</strong> 100% tied to OpenAI’s model ecosystem —<strong> no Claude, no Gemini, no alternatives</strong></li><li><strong>📈 Scaling cost risk:</strong> API costs grow linearly with usage — at massive scale, it gets expensive fast</li><li><strong>🏦 Data privacy:</strong> All customer interactions flow through OpenAI’s APIs — a dealbreaker for banks, hospitals, and regulated industries</li><li><strong>🧾 Outcome fee ambiguity:</strong> How exactly do you audit and measure “value created”? This can become a contractual grey area</li><li><strong>🏗️ Legacy complexity:</strong> For enterprises with deeply embedded legacy systems, the FDE model may not go deep enough</li></ul><h3>🧭 So, What Does This All Mean?</h3><p>OpenAI’s Deployment Company is the most credible direct challenge to the IT outsourcing model in its history. It combines the speed of a SaaS product with the hands-on delivery of a consulting firm — and prices it in a way that makes the traditional model look bloated by comparison.</p><p>But it isn’t a wholesale replacement. Not yet.</p><p>For <strong>fast-moving companies</strong> wanting rapid AI ROI with manageable complexity, OpenAI’s model is genuinely compelling. For <strong>large enterprises</strong> in regulated industries, with complex legacy infrastructure, or with data sovereignty requirements — the traditional IT firm remains the safer, more controllable bet.</p><p>The real disruption isn’t just in the cost or the speed. It’s in the <strong>pricing philosophy</strong>. OpenAI is betting that the future of enterprise IT isn’t “pay us for hours” — it’s “pay us for outcomes.” 🚀</p><p>And if that bet lands, the $250B IT outsourcing industry will never look the same again.</p><p><strong>PS: By the way, if you enjoyed reading, that’s all thanks to my creative spark! 🎨 A big thank you to Perplexity for helping me weave my ideas into this fun article — couldn’t have done it without their friendly assist. 😊</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9930567d111a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Anthropic’s Plugins Trigger the “SaaSpocalypse” ]]></title>
            <link>https://medium.com/@kannankannan18/anthropics-plugins-trigger-the-saaspocalypse-14fe24ec55e7?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/14fe24ec55e7</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[information-technology]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Wed, 04 Feb 2026 15:41:40 GMT</pubDate>
            <atom:updated>2026-02-04T15:41:40.384Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AXQ1P8F5bUy7pESHC-2X0w.png" /></figure><p>Remember when everyone said AI would help businesses, not replace them? Well, Anthropic just made everyone nervous with their latest move! On January 30, 2026, they dropped 11 new plugins for Claude Cowork, and suddenly tech stocks around the world started crashing like dominoes. We’re talking about a whopping $285 billion wiped off from the market in just one day. Let’s dive into what actually happened and why everyone’s freaking out!</p><h3>What’s This New Update Everyone’s Talking About? 🤔</h3><p>So here’s the scoop: Anthropic released something called plugins for their Claude Cowork platform. Think of plugins like apps on your phone, but for AI work assistants. They launched 11 different plugins that help Claude do specific jobs without needing much human help.</p><p>These plugins are basically bundles of skills, tools, and commands that make Claude super smart at specific tasks. The cool (or scary, depending on who you are) part is that anyone can customize them or even build their own without needing to be a coding expert. Anthropic made them open-source and put them on GitHub for everyone to access.</p><p>Here are some of the main plugins they released:</p><ul><li><strong>Legal Plugin</strong> — Reviews contracts, handles NDAs, tracks compliance</li><li><strong>Sales Plugin</strong> — Connects to CRM systems, does prospect research</li><li><strong>Marketing Plugin</strong> — Creates campaigns and content</li><li><strong>Finance &amp; Accounting Plugin</strong> — Handles financial analysis</li><li><strong>Data Analyst Plugin</strong> — Processes and analyzes data</li><li><strong>Customer Support Plugin</strong> — Manages customer queries</li><li><strong>Product Management Plugin</strong> — Helps with product workflows</li></ul><h3>The Big Problem for Tech Service Companies 😰</h3><p>Now here’s where things get messy for companies like Infosys, TCS, Wipro, and even global giants like Accenture and ServiceNow. These companies make billions by doing exactly what these plugins can now do — legal work, data analysis, customer support, financial services, you name it.</p><p>Traditional IT service companies charge clients for human hours spent on these tasks. But if an AI plugin can do the same job in minutes instead of days, why would companies pay big bucks for outsourcing? The fear is real: a Bain &amp; Company report already warned that up to 30% of tech services revenue could disappear because of AI automation.</p><p>Indian IT stocks took a massive beating on February 3, 2026:</p><ul><li>Infosys dropped 4.51%</li><li>TCS fell 3.15%</li><li>Wipro crashed nearly 6%</li><li>HCL Technologies also tumbled</li></ul><p>Together, Indian IT companies lost about Rs 2 lakh crore in market value in a single day. Ouch! 💸</p><h3>Real World Example: The Data Analyst Plugin in Action 📊</h3><p>Let me show you how this actually works with the Data Analyst Plugin — one that’s directly threatening companies like TCS and Accenture.</p><p><strong>The Old Way (Tech Service Companies like TCS/Accenture):</strong></p><p>Imagine a retail company wants to analyze their quarterly sales data across 500 stores to identify trends and optimize inventory:</p><ul><li>They hire a data analytics team from TCS or Accenture</li><li>Team size: 3–5 data analysts working on the project</li><li>Process: Data cleaning, exploratory analysis, visualization, reporting</li><li>Total time: 3–4 weeks of work</li><li>Cost: Around $50,000-$80,000 for the project (at typical consulting rates of $100–150 per hour per analyst)</li><li>Deliverables: PowerPoint presentations, Excel reports, dashboard setup</li></ul><p><strong>The New Way (Anthropic’s Data Analyst Plugin):</strong></p><p>With Claude’s Data Analyst Plugin, the same company can:</p><ul><li>Upload sales data directly to Claude Cowork</li><li>The plugin automatically cleans and processes the data</li><li>Generates insights, identifies trends, creates visualizations</li><li>Produces actionable recommendations for inventory optimization</li><li>Total time: 2–3 days with minimal human supervision</li><li>Cost: Just the Claude subscription (roughly $30–40 per user per month)</li><li>Deliverables: Interactive reports, charts, and data-driven insights ready to use</li></ul><p><strong>The Cost Impact:</strong></p><p>The difference is staggering! A $60,000 consulting project becomes a $100 AI subscription task. That’s a 99.8% cost reduction. Even if you need a human analyst to review and refine the AI’s work for a few hours, you’re still looking at maybe $5,000 total versus $60,000. This is why tech service companies are panicking — their entire pricing model just became obsolete overnight.</p><h3>Why Did Stocks Crash in the US and India? 📉</h3><p>The market reaction was swift and brutal, hitting both American and Indian companies hard.</p><p><strong>In the United States:</strong></p><p>The Goldman Sachs software basket fell 6%, and Nasdaq dropped 2.4% at peak. SaaS companies like Salesforce and ServiceNow fell around 7%, while financial services firms dropped 7% too. The big fear was what investors started calling the “SaaSpocalypse” — the potential end of traditional software service companies as we know them.</p><p>Jefferies investment bank noted that Anthropic is no longer just supplying AI models to other companies; they’re building complete workflow solutions themselves. This means foundation model companies are now competing directly with application layer companies. Why pay for specialized software when Claude can do it all with plugins?</p><p><strong>In India:</strong></p><p>Indian IT stocks took an even harder beating because these companies built their entire business model around providing skilled labor for tasks AI can now automate. They’re heavily dependent on US clients for data analysis, financial services, and business process work — exactly what these plugins target.</p><p>The loss of Rs 2 lakh crore in a single day reflected a brutal reality check. Indian IT companies were already facing slow deal pipelines and stretched valuations. When your core business is billable hours and AI can do the work in minutes, that’s an existential threat to your revenue streams.</p><h3>What Does the Future Look Like? 🔮</h3><p>Okay, let’s not get too doom-and-gloomy here. Yes, things are changing fast, but it’s not necessarily the end of the world for tech services.</p><p><strong>Short-term:</strong></p><p>More companies will experiment with AI plugins for routine tasks, and tech service firms will face pricing pressure. Hybrid models will emerge where AI handles routine work and humans handle complex cases. Job roles will shift from doing the work to managing and reviewing AI outputs. Smart companies will start retraining their workforce to focus on AI integration and oversight rather than manual execution.</p><p><strong>Long-term:</strong></p><p>Tech companies that adopt a “zero-based approach” — rebuilding their entire operations with AI at the center — could actually increase margins by 200–300 basis points according to Bain &amp; Company. New opportunities will emerge in AI implementation, customization, and integration services. The focus will shift from “doing tasks” to “solving complex problems” that AI can’t handle alone. Companies with deep industry expertise, client relationships, and ability to handle edge cases will find their niche.</p><p><strong>What Should We Take Away?</strong></p><p>The Anthropic plugin launch is a reality check: AI is changing how work gets done. Tech service companies that stick to the old billable-hours model will struggle, but those that embrace AI and focus on high-value strategic work will thrive. For tech workers, it’s time to upskill and focus on creativity, strategy, and complex problem-solving! 🚀</p><p><strong>PS: By the way, if you enjoyed reading, that’s all thanks to my creative spark! 🎨 A big thank you to Perplexity for helping me weave my ideas into this fun article — couldn’t have done it without their friendly assist. 😊</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=14fe24ec55e7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Is AI the New Dot-Com Bubble or a Tech Revolution? ]]></title>
            <link>https://medium.com/@kannankannan18/is-ai-the-new-dot-com-bubble-or-a-tech-revolution-e0f5caa3ec1f?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/e0f5caa3ec1f</guid>
            <category><![CDATA[business]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[investing]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Thu, 20 Nov 2025 17:32:34 GMT</pubDate>
            <atom:updated>2025-11-20T17:32:34.883Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/696/1*tmaX2GAY8QqwkeYmi34qxg.png" /><figcaption>Source: Bloomberg</figcaption></figure><p>A viral diagram above showing billions bouncing between Nvidia, OpenAI, Microsoft, and Oracle has sparked wild debate about a looming “AI bubble.” 😅 In this article, we’ll quickly decode <strong>what a bubble burst</strong> actually is, zoom through <strong>history’s famous crashes</strong>, and break down the <strong>Dot-com bubble</strong> in simple story style. You’ll get an easy explainer of that popular AI diagram and dive into the AI Bubble Burst theory — plus, we’ll tackle the big question: is an AI bubble burst just a juicy headline, or could it really pop? 🤔</p><h3>What Is a Bubble Burst? 🤔</h3><p>A bubble burst in technology means that a market or sector has grown super fast, with prices and valuations going way beyond what the actual value or profits can support. When reality hits, prices suddenly crash, and many companies go under or lose a lot of value. It’s like a balloon that gets filled with too much air and then pops! 🎈💥</p><ul><li>A bubble happens when investors get super excited and put lots of money into something new, even if it’s not making much profit yet.</li><li>When the excitement fades, and people realize the real value, the market crashes hard.</li></ul><h3>Famous Bubble Bursts in History 🌍</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/756/1*axXnLazNVDdTiJ7SD8Q9xw.png" /><figcaption>Bubble Bursts in the Past</figcaption></figure><h3>The Dot-com Bubble: How It Sparked &amp; Why Investors Went Crazy 🚀💸</h3><ul><li>The Dot-com Bubble started in the mid-1990s as the Internet promised huge new business potential.</li><li>Netscape’s 1995 IPO set off investment mania, sending VC money and stock prices soaring.</li><li>Funding jumped from $8.2B in 1995 to $104B in 2000; NASDAQ rocketed from 751 to 5,000+ (a 600% surge).</li><li>In 1999, NASDAQ alone grew 84% as everyone chased the “can’t-miss” internet dream.</li></ul><h3>Why Did the Dot-com Bubble Burst? 😬⚠️</h3><ul><li>The burst began when investors saw that many dot-com companies were burning cash but making almost no real profit, especially from late 1999 to early 2000.</li><li>Big names like Priceline and Webvan reported huge losses, shaking confidence and triggering heavy selling.</li><li>By March 2000, insider and institutional selling caused a domino effect, and the NASDAQ, which had passed 5,000, crashed, wiping out roughly $1 trillion in value within weeks.</li><li>Hundreds of startups such as <a href="http://pets.com/">Pets.com</a> and <a href="http://etoys.com/">eToys.com</a> went bankrupt, and over the next three years only a few survivors like Amazon and eBay made it through while most others disappeared.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/775/1*LPmF2i6670RK-iijl61X8A.png" /><figcaption><strong>NASDAQ: Before &amp; After the Downfall (Dot-com Bubble)</strong></figcaption></figure><blockquote>Fun Fact: After the bubble burst in March 2000, the NASDAQ dropped by around 78% from its peak (over 5,000) to a low near 1,140 by late 2002. It took a staggering 15 years for the NASDAQ to bounce back to its old bubble high, <strong>finally crossing 5,048 again in March 2015</strong>!</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/715/1*-mTQ7Ffa2SKlt9X4kmjH6g.png" /></figure><h3>What Is the AI Bubble Burst? 🤖💥</h3><p>The “AI Bubble Burst” is a term used to describe a possible future scenario where the current excitement and huge investments in artificial intelligence (AI) suddenly collapses.</p><p>The following diagram is the perfect snapshot of why so many market experts are worried about an “AI Bubble Burst.”</p><ul><li>The circles represent major companies and AI startups.</li><li>The size of the circle shows how valuable each company is on the market.</li><li>Thick arrows connect the companies, color-coded to show different types of relationships.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/696/1*tmaX2GAY8QqwkeYmi34qxg.png" /><figcaption>Source: Bloomberg</figcaption></figure><h3>Why This Diagram Fuels Bubble Fears ⚠️</h3><ul><li><strong>Extreme concentration:</strong> The biggest circles — Nvidia ($4.5T) and OpenAI ($500B) — dominate the map; if one stumbles, the shock can spread across the whole network.</li><li><strong>Big, complex bets:</strong> OpenAI, Nvidia, Oracle, AMD, Intel, and Microsoft are all tightly linked, with billions riding on everything going right for AI.</li><li><strong>Huge upfront spending:</strong> Deals like Nvidia’s plan to invest up to $100B in OpenAI and OpenAI’s $300B cloud deal with Oracle only pay off if AI growth continues strongly.</li><li><strong>Bubble signs:</strong> A dense web of money flows, high valuations, and a few “golden” companies at the center looks very similar to the late‑1990s dot‑com setup.</li><li><strong>Profit worries:</strong> Many AI startups still don’t have steady profits, while big costs for data centers, chips, and staff mean some are already facing pressure and falling stock prices.</li></ul><h3>🥊 Dot-Com Bubble vs AI Bubble: Fears and Reality</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/754/1*zTYvaOoulFnSEswgXOQKFw.png" /><figcaption>Dotcom Bubble vs AI Bubble</figcaption></figure><h3>So, Burst or Boom? 🎉💥</h3><p><strong>Parallels: </strong>Both phases are packed with hype, FOMO, and huge amounts of money chasing the “next big thing,” so calling AI a potential bubble isn’t totally wrong. Investors are again paying big premiums based on future dreams more than present profits.</p><p><strong>Key Differences:</strong></p><ul><li>AI is mostly led by established, cash-rich companies — not fragile startups betting everything on a dream.</li><li>There’s real, broad adoption of AI in products people use every day — unlike the 1990s web, which was just starting to catch on.</li><li>The capital spent is building cloud, chip, and data infrastructure, not just websites and advertising.</li><li>Core players like Nvidia, Microsoft, and Oracle are showing real profits and growth, not just paper valuations</li></ul><p>Experts say the AI boom is not just hype — <strong>it’s backed by real-world tech</strong> in health, finance, and much more, and strong profits from big companies. While there could be a shakeup if things don’t go as expected, the <strong>risks are centered in a few major players, not spread everywhere</strong> like in the dot-com days. So even if some parts of the bubble burst, most of the AI growth should keep going, making life and business even smarter! 😊</p><p>PS : By the way, if you enjoyed reading, that’s all thanks to my creative spark! 🎨 A big thank you to Perplexity for helping me weave my ideas into this fun article — couldn’t have done it without their friendly assist. 😊</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e0f5caa3ec1f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Buggy Update Breaks the Web : Cloudflare’s Nov 18, 2025 Outage Explained! ]]></title>
            <link>https://medium.com/@kannankannan18/buggy-update-breaks-the-web-cloudflares-nov-18-2025-outage-explained-a3989cf9019e?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/a3989cf9019e</guid>
            <category><![CDATA[cloudflare]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[web-development]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Wed, 19 Nov 2025 14:09:32 GMT</pubDate>
            <atom:updated>2025-11-19T14:22:16.414Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rvzfMesU_Hry1vIOCFpQTw.png" /></figure><p>This article takes you through Cloudflare’s major outage on November 18, 2025, when a buggy update led to big parts of the web — including X, ChatGPT, and many others — going offline for hours. You’ll learn what Cloudflare is, what went wrong, how they fixed it, and the key lessons to avoid future chaos, all in a fun and simple style with lots of real-world examples and smileys! 🌐😬😊</p><h3>Meet Cloudflare: The Friendly Web Guardian</h3><p>Cloudflare is like the invisible engine room powering a huge part of the internet, making websites faster, safer, and more reliable for millions of people every day. It’s a company that started out fighting spam, but now protects and boosts everything from tiny blogs to giant businesses.</p><p><strong>What does Cloudflare actually do?</strong></p><ul><li><strong>Speeds things up:</strong> It delivers website content from servers close to you, so pages load quickly no matter where you are.</li><li><strong>Keeps out the bad stuff:</strong> Its security features stop cyberattacks (like DDoS), block malicious bots, and prevent hackers from causing trouble.</li><li><strong>Simplifies management:</strong> It helps sites control traffic, balance loads, manage domain names, and encrypts data for safer browsing.</li><li><strong>Makes the web more stable:</strong> If one server fails, Cloudflare reroutes traffic so the site stays online.</li></ul><p><strong>Why is Cloudflare so important?</strong></p><ul><li>It sits between internet users and websites, filtering out threats and speeding up access.</li><li>Without Cloudflare, countless websites would be slower, less secure, and way more likely to crash during big traffic spikes.</li><li>As of 2025, Cloudflare protects nearly 20% of all websites in the world, handling tens of millions of requests every second.</li></ul><p><strong>Example:</strong> When you visit your favorite social media page, Cloudflare checks the connection, blocks anything risky, and loads images or messages from the nearest server — making the whole experience smooth and safe</p><h3>Cloudflare’s Nov 18 Meltdown 🚨</h3><p>On November 18, 2025, Cloudflare rolled out a “<strong>buggy update”</strong> to its system that helps spot fake internet traffic. Because of this buggy update, a key part of Cloudflare suddenly thought almost everyone trying to visit a website was suspicious — even if they were normal users.</p><p>This meant major sites like X, ChatGPT, and Perplexity stopped working for hours. Many people saw errors or couldn’t log in, just because the buggy update kept confusing Cloudflare’s software. It wasn’t a cyberattack — just a simple internal mistake that ended up causing chaos for big parts of the internet! 😬</p><h3>Technical Deep Dive into the Buggy Update! 🚨</h3><ul><li>The bot management system depends on something called a “feature configuration file.” This file is filled with traits and patterns used by machine learning to decide if a visit is legit or from a bot.</li><li>Normally, this file stays a certain size. But a change in database permissions caused the system to output duplicate lines, <strong>making the file double in size overnight</strong>.</li></ul><blockquote>Let’s look at “change in db permission” in detail:</blockquote><blockquote>Cloudflare’s Bot Management feature regularly generates a configuration file from a ClickHouse database, listing all the traits machine learning uses to spot bots. This file is refreshed every few minutes using a database query</blockquote><blockquote><strong>The change involved granting explicit permissions to access underlying tables</strong>, rather than just implicit access. This meant when a metadata query like <strong>SELECT name, type FROM system.columns WHERE table = &#39;http_requests_features&#39;</strong> was run, it started including columns from both the default and an additional schema (called r0)</blockquote><blockquote>The query didn’t filter by database name, so the returned results now had duplicate rows for each feature — one from each schema. The file generated from these results included every duplicate, ballooning in size and doubling the number of features it listed overnight.</blockquote><ul><li>The routing software that reads this file had a limit — when the file suddenly became much larger, that limit was exceeded, and the bot management module crashed for any service that depended on it.</li><li>This meant Cloudflare’s systems started to see every request (even legit human visits!) as “suspicious.” The proxy couldn’t process traffic right, so it began sending HTTP 5XX error codes — which basically tell the internet, “Sorry, something is broken, you can’t come in!”.</li><li>The new buggy update was regenerated every few minutes, so even when some servers recovered, the bad files kept sneaking back in, causing a yo-yo effect of outages for sites everywhere.</li><li>This wasn’t an attack or a hack, but an internal system bug from a rushed configuration change — proving how even small mistakes can cause big internet problems!</li></ul><h3>How did X, ChatGPT, Perplexity (and More) Go Down due to this Buggy Update? 🛑</h3><p>X, ChatGPT, Perplexity, and many other popular services went down because they all use Cloudflare to manage traffic, protect against bots, and make their websites load faster. When Cloudflare’s bot management update turned buggy, it started flagging every single request as suspicious — even if it was a real person trying to log in or use a service.</p><ul><li>The bot management crash caused Cloudflare’s servers to respond with HTTP 5XX errors, meaning all requests to these sites (like X, ChatGPT, and Perplexity) were bounced right back with a “service unavailable” message.</li><li>Because these platforms depend on Cloudflare for routing and authentication, once Cloudflare was broken, they couldn’t reach users, load web pages, or handle logins. Basically, the websites were shut off from the whole world for a while! 😬</li><li>The bug also affected other big names like Canva, Dropbox, and some banking and gaming services, showing just how much of the internet relies on Cloudflare’s tech for daily operations.</li></ul><h3>How Did Cloudflare Fix the Mess? 👨‍🔧🛠️</h3><p>Cloudflare handled the crisis with both a quick temporary fix and a deeper permanent solution to fully resolve the outage.</p><p><strong>Immediate Temporary Fixes</strong></p><ul><li>Engineers quickly stopped the creation and spread of the faulty bot management configuration file by rolling back the recent database permissions update.</li><li>They manually inserted a known-good (safe, small) version of the feature configuration file, replacing the bad version across Cloudflare’s global network.</li><li>Emergency bypasses were activated for essential systems like Workers KV and Cloudflare Access, allowing critical services to resume even before the core issue was fully fixed.</li><li>Teams continuously monitored the system and, as soon as the old-safe file was confirmed working, rushed to deploy it on every node, speeding up the global recovery.</li></ul><p><strong>Permanent Solutions</strong></p><ul><li>Cloudflare permanently fixed the bug by correcting the underlying query and permissions in their database. The update now strictly targets only the intended data table/schema, avoiding duplicates.</li><li>They tightened monitoring on configuration file sizes, adding real-time alerts for abnormal changes and extra guardrails to stop bad data before distribution.</li><li>All future critical updates now require thorough live testing and have “kill switch” rollout controls to stop problem changes before they can cause another mass outage.</li><li>The company reviewed and improved its error-handling for other vital parts of their proxy, ensuring one glitch can’t so easily break so many platforms at once.</li></ul><p><strong>Time Taken to Resolve ⏲️</strong></p><ul><li>The most severe part of the outage lasted about three hours: First errors started at ~11:20 UTC, with core network traffic largely restored by 14:30 UTC.</li><li>Final, lingering issues were resolved and all systems returned to normal by 17:06 UTC, making the full timeline about six hours from the start to complete stabilization.</li></ul><h3>Conclusion: Building a Stronger (and Safer!) Internet 🚦</h3><p>The Cloudflare outage showed how even a small change can ripple into a huge web disruption. Here’s how to be better prepared next time:</p><ul><li><strong>Test for Weird Scenarios:</strong> Always stress-test changes, not just under normal conditions but also under rare or extreme ones.</li><li><strong>Have Fast Rollbacks:</strong> Quick emergency “off switches” for new updates help limit damage when things go wrong.</li><li><strong>Build Redundancies:</strong> Use backups or multi-provider setups, so a single failure doesn’t halt your service.</li><li><strong>Know Your Dependencies:</strong> Map out what connects to what, because a problem in one layer (like bot management) can break everything above it.</li><li><strong>Turn Monitoring into Action:</strong> Set alerts that prompt immediate investigation — not just log issues quietly.</li><li><strong>Plan for Outages:</strong> Practice how you’ll respond if a big part of your tech stack fails, so you aren’t caught off guard.</li></ul><p>Being proactive and resilient means the next hiccup won’t take the whole web down. Let’s make outages shorter, and our digital world stronger! 🌍✨</p><p>PS : By the way, if you enjoyed reading, that’s all thanks to my creative spark! 🎨 A big thank you to Perplexity for helping me weave my ideas into this fun article — couldn’t have done it without their friendly assist. 😊</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a3989cf9019e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why Gold Shines Brighter (and Cheaper) in Dubai than in India ✨]]></title>
            <link>https://medium.com/@kannankannan18/why-gold-shines-brighter-and-cheaper-in-dubai-than-in-india-6c75c0a468a8?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/6c75c0a468a8</guid>
            <category><![CDATA[gold]]></category>
            <category><![CDATA[articles]]></category>
            <category><![CDATA[trade]]></category>
            <category><![CDATA[dubai]]></category>
            <category><![CDATA[india]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Fri, 24 Oct 2025 03:39:04 GMT</pubDate>
            <atom:updated>2025-10-24T03:39:04.566Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZX5jZZRrGIA4N0IWPTH5WA.png" /></figure><p>Ever wondered why everyone talks about buying gold in Dubai? Whether it’s friends returning from vacations flaunting glittering jewellery or traders making bulk purchases, the golden price gap between Dubai and India is real — and fascinating! Let’s take a dive into the story behind it — prices, policies, and even what the future might hold. 🌟</p><h3>The Real “Gold Standard”: Where Prices Begin 🌏</h3><p>Every gram of gold in Dubai and India starts from one single source — the <strong>international spot price</strong>, determined by the <strong>London Bullion Market Association (LBMA)</strong>.</p><ul><li>LBMA conducts <strong>two daily auctions</strong> (10:30 AM &amp; 3:00 PM London time) through the <strong>ICE Benchmark Administration</strong>.</li><li>Global banks like <strong>HSBC, JP Morgan, Citi, and Bank of China</strong> participate to find the price where global supply meets demand — that becomes the <strong>official spot rate</strong>.</li><li>As of <strong>October 24, 2025</strong>, the LBMA gold price sits near <strong>$4,112 per ounce</strong>, or <strong>$1,322 per 10 grams</strong>.</li><li>At today’s <strong>USD-INR rate of ₹88.02</strong>, this converts to a <strong>base price of about ₹116,336 per 10 grams</strong> (or ₹11,633 per gram).</li></ul><p>This is the <strong>global base price</strong> — before countries add their own import duties, taxes, or retailer margins.</p><h3>How the Base Price Turns Pricier in India 💸</h3><p>While both India and Dubai start from the same LBMA-linked global rate, the final retail price depends on local taxes and trade policies:</p><ul><li><strong>Dubai:</strong> <strong>No import duty</strong>, 5% VAT (refundable for tourists).</li><li><strong>India:</strong> <strong>6% import duty (down from 15% earlier)</strong>, 3% GST + 5% GST on making charges.</li></ul><p>That’s why Dubai’s gold stays 6–7% cheaper than India’s at any given time</p><h3>Why Dubai Keeps Gold Cheaper 💎</h3><ul><li><strong>Duty-Free Imports:</strong> Gold enters free of customs tariffs.</li><li><strong>Refundable VAT:</strong> Tourists claim it back before departure.</li><li><strong>Gold Souk Competition:</strong> Hundreds of shops keep margins razor-thin.</li><li><strong>Purity Certifications:</strong> Every product is Hallmarked under <em>Bareeq Certification</em>.</li></ul><p>Dubai purposefully built its reputation as “The City of Gold” by keeping policies investor‑friendly and prices internationally competitive.</p><h3>Why India’s Gold Costs More 🇮🇳</h3><ul><li><strong>Import Dependency:</strong> India imports about 800 tonnes of gold every year, largely through dollar purchases.</li><li><strong>Economic Management:</strong> Import duties control excessive gold imports and help maintain foreign exchange reserves.</li><li><strong>Revenue Generation:</strong> Taxes and duties add billions to state revenues, making policy change unlikely</li></ul><h3>Example: Today’s 24K Gold Price Breakdown (As of October 24, 2025) 💰</h3><p>This example shows how the same gram of 24K gold is priced in Dubai versus India.</p><p><strong>Global Base Price (LBMA)</strong></p><ul><li><strong>Spot Price:</strong> $1,322 per 10 grams</li><li><strong>Exchange Rate:</strong> ₹88.02 per USD</li><li><strong>Base Value:</strong> ₹116,336 per 10 grams (₹11,633 per gram)</li></ul><p><strong>Per‑Gram Comparison (Table View) 📊</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zsts-OMaM3OOeIwd2kkmHA.png" /><figcaption>Gold Price Derivation (Dubai vs India)</figcaption></figure><p>Dubai’s rate is nearly <strong>₹1,000 cheaper per gram</strong>, or about <strong>₹10,000 per 10 grams</strong> less than India — just because of policy differences!</p><h3>Why Dubai Doesn’t Need Import Duties ? 🌟</h3><p>Unlike India, Dubai doesn’t depend on gold import duties to run its economy. The UAE’s revenue sources are deeply diversified:</p><ul><li><strong>Non‑oil sectors account for over 75% of UAE’s GDP</strong> in 2025 <strong>(UAE Ministry of Economy, 2025)</strong>, driven by finance, tourism, logistics, and technology .</li><li>The country’s gold and commodities trade touched nearly <strong>USD 170 billion in 2024</strong>, with exports alone worth <strong>over USD 53 billion</strong> .</li><li>Major hubs like the <strong>Dubai Multi Commodities Centre (DMCC)</strong> and the <strong>Dubai Gold &amp; Commodities Exchange (DGCX)</strong> generate massive revenues through licenses, trading fees, and value‑added services instead of taxation.</li></ul><p>This means Dubai <strong><em>doesn’t</em> <em>need</em></strong> to tax imports — other thriving industries like tourism, real estate, and banking fund its economy robustly.</p><p>India, on the other hand, still <strong>leans heavily on taxation</strong> to support its large budget. Easing import duties would require India to deepen alternate revenue streams such as digital exports, services, and manufacturing before it can afford to go low‑tax like Dubai.</p><h3>The Golden Bottom Line 🌠</h3><p>The same ounce of gold starts at the same global price — but ends differently based on tax philosophies:</p><ul><li><strong>Dubai = Global Rates + Tiny Markup (Accessible &amp; Investor Friendly)</strong></li><li><strong>India = Global Rates + Taxes &amp; Duties (Economy‑Protective)</strong></li></ul><p>So the next time someone returns from Dubai with a sparkly souvenir, you’ll know exactly why it shines a bit brighter (and cost way less). 💛✨</p><p><strong>PS : By the way, if you enjoyed reading, that’s all thanks to my creative spark! 🎨 A big thank you to Perplexity for helping me weave my ideas into this fun article — couldn’t have done it without their friendly assist. 😊</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6c75c0a468a8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From FOMO to Fundamentals: Navigating Silver’s 2025 ETF Rush]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@kannankannan18/from-fomo-to-fundamentals-navigating-silvers-2025-etf-rush-22624302996d?source=rss-c6778e363b04------2"><img src="https://cdn-images-1.medium.com/max/600/1*tgUsBBrmahtS9mj9Ziq1Qg.avif" width="600"></a></p><p class="medium-feed-snippet">Guess What? Silver&#x2019;s on Fire! &#x1F31F;</p><p class="medium-feed-link"><a href="https://medium.com/@kannankannan18/from-fomo-to-fundamentals-navigating-silvers-2025-etf-rush-22624302996d?source=rss-c6778e363b04------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@kannankannan18/from-fomo-to-fundamentals-navigating-silvers-2025-etf-rush-22624302996d?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/22624302996d</guid>
            <category><![CDATA[investment]]></category>
            <category><![CDATA[money-making-ideas]]></category>
            <category><![CDATA[gold]]></category>
            <category><![CDATA[etf]]></category>
            <category><![CDATA[silver]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Fri, 17 Oct 2025 16:25:25 GMT</pubDate>
            <atom:updated>2025-10-17T16:25:25.552Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Client Side Pagination vs Server Side Pagination]]></title>
            <link>https://medium.com/@kannankannan18/client-side-pagination-vs-server-side-pagination-576a8f57257d?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/576a8f57257d</guid>
            <category><![CDATA[web-app-development]]></category>
            <category><![CDATA[database]]></category>
            <category><![CDATA[pagination]]></category>
            <category><![CDATA[front-end-development]]></category>
            <category><![CDATA[backend-development]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Sun, 19 Mar 2023 12:13:03 GMT</pubDate>
            <atom:updated>2023-03-22T15:11:08.600Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YqjgFPgOSP34BH24SmhfVQ.png" /><figcaption>Client vs Server side Pagination</figcaption></figure><p>Consider that we are creating a website for your HR/TA (Talent Acquisition) team to assist them by streamlining the hiring process. Take into account that about 10,000 candidates have registered. They must be able to use the app to do the following tasks as HR :</p><ul><li>see the candidate’s list</li><li>ability to search for the required candidate</li><li>ability to filter candidates based on certain fields</li><li>ability to sort the candidate’s data</li></ul><p>It is our duty as developers or architects to create these functionalities in a user-friendly and effective way. Understanding how to handle the data would be essential to achieving that.</p><p>In a typical web application scenario, this is how a data flow happens :</p><ol><li>Frontend systems would make an API request call to the Backend systems asking for the Data</li><li>The Backend systems, on receiving the request, will contact the Database using a query to fetch the data.</li><li>The Database is where the actual data will be stored. On receiving the request from the Backend system, the DB would send the necessary data to the Backend</li><li>The Backend sends back the data to the Frontend as a Response</li><li>The Frontend systems will display the data as per user needs</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rqb0m0b-DdmpW70a9ilGRA.png" /><figcaption>DATA FLOW DIAGRAM</figcaption></figure><p>Out of the 5 steps mentioned above, this document will be focusing on the 4th step. i.e. “<strong>How do we send the data from the server(Backend systems) ?”</strong></p><p>I want to emphasise the term “Pagination” used here before getting into the real matter. Everyone (including me) would have immediately thought of the page numbers that appear beneath tables or records lists when hearing the phrase pagination. That is one interpretation of it, however in this context, “pagination” refers to “the method by which data is transmitted from the server to the client.”</p><p><strong><em>We shall be discussing 3 types of Pagination types :</em></strong></p><ol><li><strong><em>Server Side Pagination</em></strong></li><li><strong><em>Client Side Pagination</em></strong></li><li><strong><em>A combination of both Client and Server Side</em></strong></li></ol><p><strong>The pagination type needs to be chosen based on the answers to the following requirement questions -</strong></p><ol><li><strong>Will our approach handle complex data and complex data fetching logic?</strong></li><li><strong>Will our approach work efficiently for both smaller and huge data sets?</strong></li><li><strong>Will our approach work effectively for users who use Searching and Filtering options regularly?</strong></li></ol><h3>Server Side Pagination</h3><p>When the backend/server only returns a portion of the actual data requested by the client, this is known as server-side pagination. If you created a query to retrieve all candidates, and there are 10,000 candidates in total, only 100 records will be returned to the front end. To obtain the subsequent batch of records, another API request is issued.</p><blockquote>Server Side Pagination is like ordering only what’s necessary at that point in time and then making subsequent orders if needed.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OCLtaska50pebc4fmgRJrw.png" /><figcaption>SERVER-SIDE PAGINATION</figcaption></figure><p><strong>Pros</strong></p><ul><li>the initial load will be quick as only fewer data is being returned from the server</li><li>fewer chances of loading inconsistent data</li><li>performance shall be stable even if the application and data grow bigger</li><li>can handle 100s to millions of records without creating any impact on performance</li><li>loads and saves only specific/required data in the client</li><li>Search, Filter and Sorting will be faster for a large number of records as we return only the matching records, whereas in CSP we need to do these on FE and against the large data set which could result in slow processing as the browser can’t handle the heavy load as the server could</li><li>works better on complex datasets</li></ul><p><strong>Cons</strong></p><ul><li>lots of API calls need to be made to fetch data</li><li>Every page load takes equal time in fetching the data whereas with CSP only the initial load is time-consuming</li><li>we need to make an additional query to fetch the total number of records</li></ul><p><strong>You can use Server Side Pagination, when</strong></p><ul><li>when the data size is large</li><li>Your app needs to display consistent data always (eg: stock market data)</li><li>you want to have a quick initial load and also subsequent loads happening within a specific timeframe</li></ul><h3>Client Side Pagination</h3><p>When the backend/server returns all the data a client has requested, this is known as client-side pagination. If you ran a query to retrieve all of the applicants and there were 10,000 candidates in total, all 10,000 records would be returned at once to the front end.</p><blockquote>Client Side Pagination is like ordering all the food at once in a restaurant. From soup to dessert, everything is served at once. You can even start with Dessert and end with soup.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*96y8t6lNNaXdcg9K6rL0uQ.png" /><figcaption>CLIENT-SIDE PAGINATION</figcaption></figure><p><strong>Pros</strong></p><ul><li>though the initial load would be slow, subsequent page renders will be very fast as the data is already available on the client</li><li>Searching, Filtering, and Sorting operations will be very fast for less number of data sets (Eg: 1000 records)</li><li>Only 1 API call is made to fetch the data and no other API calls are required</li></ul><p><strong>Cons</strong></p><ul><li>initial load time will be longer and as the record size increases, the load time increases too</li><li>Data inconsistency will be there as we are fetching all the records at one time. In case there was an update in the backend unless we refresh the page, the change isn’t going to be reflected in the UI</li><li>all the data will be available on Client which increases the load and also it’s not certain that every user will need all those data</li><li>the performance will be impacted as data size increases</li><li>Searching, Sorting, and Filtering on clients becomes a time-consuming and inefficient process when the data size increases</li></ul><p><strong>You can use Client Side Pagination, when</strong></p><ul><li>when the data size is smaller</li><li>Your app doesn’t need to display consistent data always</li><li>Your app can take some time for the initial load but needs to be very faster in the subsequent display of data</li></ul><h3><strong>Combination of Client and Server Side Pagination</strong></h3><p>We can opt for a combination of both Client and Server side Pagination. Picking the best from both Client and Server Side pagination, this approach will solve all the required questions mentioned at the start.</p><p>To understand the approach in detail, let’s take some sample use cases and write the process to be followed for that.</p><p><strong>Sample use case 1</strong> — Let’s consider a data set that’s containing <strong>10,000 records.</strong> Let’s assume that we are planning to display 50 records at a time on a single page. So, we shall be having a total of 200 pages.</p><p><strong>Process :</strong></p><ol><li>Initially fetch 500 records from the API (Client Side Pagination)</li></ol><ul><li>Fetching 500 records will be a quick API call</li><li>This will have data for the first 10 pages</li><li>Users can traverse from Pages 1 to 10 very quickly as the load time will be very fast since the data is available on the client</li></ul><p>2. When the user clicks on any pages after 10, fetch the next 500 records from the API (Server Side Pagination)</p><ul><li>Not many users will traverse through the page by page as we have the search and filter functionality, hence even if the load time is a little longer while fetching new records, the user has less chance of experiencing it</li></ul><p>3. Searching and Filtering will happen on the server side as the data size is more. The final results based on Search and Filter Operation are returned and displayed</p><p><strong>Sample use case 2</strong> — Let’s consider a data set that’s containing <strong>1,000 records.</strong> Let’s assume that we are planning to display 50 records at a time on a single page. So, we shall be having a total of 20 pages.</p><p><strong>Process :</strong></p><ol><li>Same as a use case 1</li><li>Same as a use case 1</li><li>Searching and Filtering can be implemented in 2 ways -</li></ol><ul><li>Searching and Filtering will happen on the client side as the data size is less. Based on the total number of data, we can write a logic to perform a search and filter on the client side itself</li><li>Same as a use case 1</li></ul><p><strong>Pros</strong></p><ul><li>the load time shall be faster while traversing between the pages within those 500 records</li><li>reducing the number of API calls</li><li>can build flexible logic to use Client side Searching, Sorting and Filtering for lesser records and use Server side Searching, Sorting, Filter for larger records</li><li>The initial load will be faster</li><li>chances of data inconsistency are less when compared to CSP and more when compared to SSP</li><li>avoid storing all the data in the client</li></ul><p><strong>Cons</strong></p><ul><li>Data inconsistency is more when compared to SSP and less when compared to CSP</li><li>Write logic to handle both the server side and client side way of the approach to search, sort and filter functionality</li></ul><h3>Conclusion</h3><p>Each of the three pagination techniques has benefits and drawbacks, therefore, choosing the best one will largely depend on the application and business use case.</p><p>I appreciate you taking the time to read this. I’ll hopefully see you with the next one soon.</p><p><strong>Reference Links :</strong></p><ul><li><a href="https://buildthis.com/client-side-vs-server-side-pagination/">https://buildthis.com/client-side-vs-server-side-pagination/</a></li><li><a href="https://betterprogramming.pub/introduction-to-client-side-and-server-side-pagination-in-typescript-9edae72f94a0">https://betterprogramming.pub/introduction-to-client-side-and-server-side-pagination-in-typescript-9edae72f94a0</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=576a8f57257d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The What, Why and How aspect of “Scalability” in Web Application]]></title>
            <link>https://medium.com/@kannankannan18/the-what-why-and-how-aspect-of-scalability-in-web-application-3d9497730e2e?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/3d9497730e2e</guid>
            <category><![CDATA[scaling]]></category>
            <category><![CDATA[application-development]]></category>
            <category><![CDATA[scalability]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[scaleup]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Sat, 26 Feb 2022 05:31:06 GMT</pubDate>
            <atom:updated>2022-02-26T05:31:06.814Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/724/1*ePBG0pU38AWSpsyLGCP9oA.jpeg" /></figure><h3>What is “Scalability” in Web Application ?</h3><p>It is the ability of a system to handle sudden changes in workload without negatively impacting performance.</p><h3>Scalable Application definition in layman’s term</h3><p>An application that’s built to handle any number of users as it has the capability to expand(scale up) and contract(scale down) itself based on the user load that it receives. And remember, the app’s performance should be the same whether there are one user or 1000.</p><h3>Why we need to think about Scalability aspect ?</h3><blockquote>Google found that 61% of users won’t give an app a second chance if they had a bad first experience. Nearly 40% go straight to a competitor’s product instead.</blockquote><p>An application that’s built without planning the scalability will have the following drawbacks -</p><ul><li>Increase in number of users or workloads will make the application slow down and in turn affects the performance</li><li>Adding more features to it becomes an issue as it was designed to handle only certain load</li><li>If an application crashes because of this then it loses the trust that user’s had thereby making it unreliable</li></ul><h3>What helps an application in getting the “Scalable Application” badge ?</h3><ul><li><strong>Availability:</strong> The system should be always available. Uptime percentage has the most immediate effect on customer experience.</li><li><strong>Performance:</strong> The system must maintain a high level of performance even under heavy loads.</li><li><strong>Reliability:</strong> The system must accurately store, retrieve, and edit data under any possible scenario. It should return the most current data on request without defaulting to old or inaccurate data or failing to record new data.</li></ul><h3>Factors influencing scalability</h3><p>All the following factors must focus on one question in common — “Is this the right choice for my application ?” -</p><ul><li><strong>Architecture</strong> — the end-end architecture of whole application</li><li><strong>Framework</strong> — the framework that we opt to use</li><li><strong>Code</strong> — having a clean code design</li><li><strong>Load Testing</strong> — helps overcome weak sides of the application and guarantee smooth performance</li><li><strong>Hardware</strong> — analysing the hardware requirements and choosing the necessary ones</li><li><strong>Third Party Services </strong>— carefully analysing and choosing only those services that wouldn’t show impromptu behaviour</li></ul><h3>Types of Scaling</h3><p><strong>Scale Up (Vertical Scaling)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/680/1*hbGIYtuxRqudi7vTP76jDg.png" /></figure><ul><li>It talks about the ability to replace the components of the existing system with new powerful components</li><li>adding more power — CPU power, RAM, disk space, or bandwidth, existing infrastructure. Eg : Say your app is deployed in machine that runs with 4GB ram. We scale up by increasing it to 8GB ram</li><li>Scaling up offers better performance than scaling out. Everything is contained in one place, allowing for faster returns and less vulnerability.</li><li>This approach can also be applied to web servers, database servers, and other parts of your system</li><li>Advantages — low initial cost, data consistency</li><li>Disadvantages — Single point of failure, H/W and S/W limitation, Downtime</li></ul><p><strong>Scale Out (Horizontal Scaling)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1018/1*JXAguxv5Smvy3sLrkyGULQ.png" /></figure><ul><li>adding additional nodes or machines to your infrastructure</li><li>by scaling out, you share the processing power and load balancing across multiple machines.</li><li>Initial cost tends to be higher but it becomes more efficient after certain point</li><li>Advantages — Resilience &amp; Redundancy, Easy to scale</li><li>Disadvantages — Additional Components, network communication</li></ul><p><strong>Scale side-by-side (Diagonal Scaling)</strong></p><ul><li>It is a mixture of both Horizontal and Vertical scalability where the resources are added both vertically and horizontally.</li></ul><p>The Front End, Back End and DB are the major components of any Web Application. If you create a project with a scalable web application architecture, ensure correct scaling of all the available components.</p><p>Choosing the right type of scaling solely depends on having a good understanding of app’s functionality and it’s possible future enhancements. But in case if you feel, you don’t have a clear picture of the former, it’s good to opt for Horizontal scaling as you always would have an option of scaling down without being hurt much on infra and cost aspect.</p><h3>Some of the best practices for incorporating Scalability</h3><p><strong>Load Balancing Software</strong></p><p>It helps to spread the workload across servers to ensure no single server gets overwhelmed.</p><p><strong>Caching</strong></p><p>Use caching to max possible extent. This helps in saving several “non-required” API call, DB hits etc. which helps in improving the performance</p><p><strong>Try opting to Cloud Storage</strong></p><p>All the major cloud service providers have made it so easy for their customers to deploy an application in cloud without letting them worry about any security or maintenance concerns. The Cloud Providers has an internal mechanism in handling the scaling and that we don’t have to worry on implementing it manually.</p><p><strong>Use a scalable database</strong></p><p>NoSQL databases tend to be more scalable than SQL.</p><p>To summarise, if you are done with Requirement Analysis phase and stepping into Design phase of Application Development, do find out the answers to the following questions before you step into Development :</p><ol><li>How can we make our application a scalable one ?</li><li>What factors we must consider in building a scalable application ?</li><li>Why type of scalability suits our application the most ?</li><li>How do we scale the hardware in case if we face severe rise of user load in future ?</li><li>What best practices we could follow in helping the application give out it’s best performance always ?</li></ol><p>I hope you found this article useful. Will meet you with the next one soon 🙂</p><p><strong>References :</strong></p><ol><li><a href="https://gearheart.io/articles/how-build-scalable-web-applications/">https://gearheart.io/articles/how-build-scalable-web-applications/</a></li><li><a href="https://justcoded.com/blog/how-to-make-a-successful-scalable-web-application/">https://justcoded.com/blog/how-to-make-a-successful-scalable-web-application/</a></li><li><a href="https://www.conceptatech.com/blog/importance-of-scalability-in-software-design">https://www.conceptatech.com/blog/importance-of-scalability-in-software-design</a></li><li><a href="https://www.conceptatech.com/blog/dos-donts-designing-scalable-architecture">https://www.conceptatech.com/blog/dos-donts-designing-scalable-architecture</a></li><li><a href="https://blog.codingoutloud.com/2010/07/27/three-types-of-scaling-in-the-cloud-scale-up-scale-out-and-now-scale-side-by-side-with-juxtaposition-scaling/">https://blog.codingoutloud.com/2010/07/27/three-types-of-scaling-in-the-cloud-scale-up-scale-out-and-now-scale-side-by-side-with-juxtaposition-scaling/</a></li><li><a href="https://touchstonesecurity.com/horizontal-vs-vertical-scaling-what-you-need-to-know/">https://touchstonesecurity.com/horizontal-vs-vertical-scaling-what-you-need-to-know/</a></li><li><a href="https://www.freecodecamp.org/news/understanding-database-scaling-patterns/">https://www.freecodecamp.org/news/understanding-database-scaling-patterns/</a></li><li><a href="https://intervision.com/5-best-practices-database-scaling/">https://intervision.com/5-best-practices-database-scaling/</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3d9497730e2e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Create React App without Create React App]]></title>
            <link>https://medium.com/bitsrc/create-react-app-without-create-react-app-b0a5806a92?source=rss-c6778e363b04------2</link>
            <guid isPermaLink="false">https://medium.com/p/b0a5806a92</guid>
            <category><![CDATA[webpack]]></category>
            <category><![CDATA[react]]></category>
            <category><![CDATA[reactjs]]></category>
            <category><![CDATA[babel]]></category>
            <category><![CDATA[javascript]]></category>
            <dc:creator><![CDATA[Kannan Nagasamy]]></dc:creator>
            <pubDate>Wed, 16 Feb 2022 16:42:25 GMT</pubDate>
            <atom:updated>2025-03-04T17:26:33.157Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*12Xgy-DmusIsxKVyJSZFOQ.png" /></figure><p>This article talks about the process of creating react app without using any libraries or frameworks such as “create-react-app” , “NextJS” etc.</p><h4>Prerequisites concepts to know</h4><ul><li><strong>Webpack — </strong>helps in bundling our code into one single file</li><li><strong>Babel</strong> — used to convert ECMAScript 2015+ (ES6+) code into a backwards compatible version of JavaScript that can be run by older JavaScript engines.</li><li><strong>Node.js</strong> — installing node, creating package.json and installing node modules using npm</li></ul><h4>This article will help us to understand</h4><ul><li>How webpack and babel work in a more practical sense;</li><li>The actual starting to ending flow of building the React app;</li><li>How development and production build is being setup and its significance;</li><li>Setting up the server details thats required;</li><li>Writing webpack and babel config files and understanding the logic that exists there;</li><li>How we can configure client-side and server-side rendering;</li><li>Understanding how HMR is works in React.</li></ul><h4>Source code</h4><ul><li><em>Repository — </em><a href="https://github.com/kannanagasamy/react-app-without-cra"><em>https://github.com/kannanagasamy/react-app-without-cra</em></a></li><li><em>Branch — main</em></li></ul><p><strong>Read More:</strong></p><p><a href="https://blog.bitsrc.io/6-best-ways-to-create-a-new-react-application-57b17e5d331a">7 Best Ways to Create a New React Application</a></p><figure><a href="https://bit.cloud/"><img alt="Build in AI speed — Compose enterprise-grade applications, features, and components" src="https://cdn-images-1.medium.com/max/1024/1*yX0vq3scSudyT61IuWFuEg.png" /></a><figcaption>Build in AI speed — Compose enterprise-grade applications, features, and components</figcaption></figure><h3>Step by Step</h3><h4>1. Make sure node is installed in your system</h4><p>Install Node.js in your system and make sure its installed by typing node -v in your terminal.</p><h4>2. Create project folder and package.json</h4><p>Create a project folder of any name and navigate to the folder and then use npm init to create a <strong>package.json </strong>file inside the folder. Navigate to the folder.</p><h4>3. Install webpack dependencies</h4><pre>npm i --save-dev webpack webpack-cli webpack-dev-server</pre><ul><li><strong>webpack</strong> — Will allow us to bundle all of our code into a final build</li><li><strong>webpack-cli — </strong>CLI tool for providing a flexible set of commands for developers to increase speed when setting up a custom webpack project. If you’re using webpack v4 or later and want to call webpack from the command line, you need this</li><li><strong>webpack-dev-server — </strong>Webpack dev server is a mini Node.js Express <a href="http://server.it/">server.It</a> uses a library called SockJS to emulate a web socket. Will enable us to create a localhost dev environment</li></ul><h4>4. Install the Babel dependencies</h4><pre>npm i --save-dev babel-loader @babel/preset-env @babel/core <br>@babel/plugin-transform-runtime <br>@babel/preset-react <br>@babel/eslint-parser <br>@babel/runtime<br>@babel/cli</pre><ul><li><strong>babel-loader — </strong>allows transpiling JavaScript files using babel and webpack. exposes a loader-builder utility that allows users to add custom handling of Babel’s configuration for each file that it processes.</li><li><strong>@babel/preset-env — </strong>allows you to use the latest JavaScript without needing to micromanage which syntax transforms (and optionally, browser polyfills) are needed by your target environment(s). This both makes your life easier and JavaScript bundles smaller!</li><li><strong>@babel/core — </strong>core package</li><li><strong>@babel/plugin-transform-runtime — </strong>A plugin that enables the re-use of Babel’s injected helper code to save on codesize.</li><li><strong>@babel/preset-react — </strong>use react preset when we are using Reactjs. Helps in converting html files to react based file</li><li><strong>babel-eslint — </strong>parser that allows ESLint to run on source code that is transformed by Babel</li><li><strong>@babel/runtime — </strong>package that contains a polyfill and many other things that Babel can reference.</li><li><strong>@babel/cli — </strong>command line interface to use babel</li></ul><h4>5. Install required Linters and path</h4><pre>npm i --save-dev eslint eslint-config-airbnb-base <br>eslint-plugin-jest <br>eslint-config-prettier<br>path</pre><h4>6. Install react and react-dom</h4><pre>npm i react react-dom</pre><h4>7. Create index.html file</h4><p>Create folder called “<strong>public”</strong> in the root of the project. Inside that, create <strong>index.html</strong>.</p><h4>8. Create App.js file</h4><p>Create src folder and inside that create a file called <strong>App.js</strong>. Add the following code to it:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/0848e27e480c9b06eed83f45c40d4470/href">https://medium.com/media/0848e27e480c9b06eed83f45c40d4470/href</a></iframe><h4>8. Create index.js file</h4><p>Create an <strong>index.js</strong> file at the root of project or anywhere you wish to have. This will act as entry point for our webpack.</p><p>Add the following code to it:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/06d0d46ca1e41f614926d655284624ca/href">https://medium.com/media/06d0d46ca1e41f614926d655284624ca/href</a></iframe><h4>9. Create webpack.config.js file</h4><p>Create a file called <strong>webpack.config.js</strong> at the root of project and add the following code. On a higher note, this file contains configs that takes care of bundling the files into one single file and setting up the dev server.</p><p>The comments in the code helps us understand what each line does:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/8b9bc6a032d2903ca2ec6e4502e0b7e9/href">https://medium.com/media/8b9bc6a032d2903ca2ec6e4502e0b7e9/href</a></iframe><h4>10. Create .babelrc file</h4><p>Create a file called .babelrc at the root and add the following code.</p><p>This is the configuration file for Babel, and you’ll use it to tell babel to use the plugin and presets defined inside.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d841f1aa60bb578f6772522cb323e41a/href">https://medium.com/media/d841f1aa60bb578f6772522cb323e41a/href</a></iframe><h4>11. Update package.json file</h4><p>Add the start and build scripts to it — <em>line number 7 and 8.</em></p><ul><li>The <strong>start</strong> script asks to run the webpack-dev-server in our current project at port 9500, from the public folder.</li><li>The build command asks us to <strong>build</strong> this package in <strong>main.js</strong> file. It actually runs all logic in <strong>webpack.config.js</strong> file.</li></ul><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/76875f8420f492befa322b1aad7ca752/href">https://medium.com/media/76875f8420f492befa322b1aad7ca752/href</a></iframe><h4>12. Final project folder structure has to be like this</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/516/1*rKVs9s3qtTTuZc1tmJ3K8w.png" /></figure><h4>13. Run “npm run build”</h4><ul><li>Once added the above code, hit npm run build. it generates <strong>main.js</strong> file in our <strong>public</strong> folder. The file is actually <strong>over 1 MB</strong> in size. This is our development build.</li></ul><h4>14. Run “npm start”</h4><p>Start the application by giving npm start from terminal. This will start our dev server.</p><p>The final code can be found in the repo link I shared above.</p><blockquote><em>💡</em><strong><em>Note:</em></strong><em> If you wish to share and reuse your React components across multiple projects, you can use an open-source toolchain like </em><a href="https://bit.cloud/"><strong><em>Bit</em></strong></a><em> which lets you create reusable components. You can develop components in isolation and publish them as standalone entities that can be searched, tested, and installed across projects. Find out more </em><a href="https://bit.dev/blog/how-to-reuse-react-components-across-your-projects-l4pz83f4/"><strong><em>here</em></strong></a><em>, and </em><a href="https://bit.dev/blog/component-driven-microservices-with-nodejs-and-bit-l64shurc/"><strong><em>here</em></strong></a><em>.</em></blockquote><p><a href="https://bit.dev/blog/how-to-create-a-composable-react-app-with-bit-l7ejpfhc/">How to Create a Composable React App with Bit</a></p><h3>Other Key Findings</h3><h4>Changing the build to production</h4><ul><li>Now we can try changing it to production build. For this you need to make the following change to <strong>webpack.config.js</strong> file.</li></ul><pre>mode: &quot;production&quot;</pre><ul><li>now running npm run build will create <strong>main.js</strong> file again but the size will be very less (&lt;200kb).</li><li>With the optimization from <strong>1000 KB</strong> to <strong>200 KB</strong>, we might want to use production build always. But, we should use development mode while doing development because the <strong>hot reloading is faster in development mode.</strong></li></ul><h4>Hot Module Replacement</h4><ul><li>HMR is handled by webpack-dev-server. We can use HMR without page load option too. Setting the required options helps greatly in performance aspect.</li><li>Check the below code snippet for various scenarios:</li></ul><pre>//If you want to use HMR but no live reload then use the below config in webpack.config.js<br>devServer: {<br>        hot: true ,<br>        liveReload:false<br>    }<br><br>//If you want don&#39;t want to use HMR but want to use live reload then,<br>devServer: {<br>        hot: false ,<br>        liveReload: true<br>    },<br><br>//If you want don&#39;t want to use live reload then,<br>devServer: {<br>        hot: false , //this is mandatory to be set to false for this<br>        liveReload: false<br>    },</pre><h4>References</h4><ol><li>HMR with webpack — <a href="https://webpack.js.org/guides/hot-module-replacement/">https://webpack.js.org/guides/hot-module-replacement/</a></li><li>Different ways to reduce the bundle size — <a href="https://blog.jakoblind.no/3-ways-to-reduce-webpack-bundle-size/">https://blog.jakoblind.no/3-ways-to-reduce-webpack-bundle-size/</a></li><li>Understanding devserver and working in details — <a href="https://webpack.js.org/configuration/dev-server/#devserverlivereload">https://webpack.js.org/configuration/dev-server/#devserverlivereload</a></li><li>Minimizing the bundle for production — <a href="https://webpack.js.org/plugins/mini-css-extract-plugin/#minimizing-for-production">https://webpack.js.org/plugins/mini-css-extract-plugin/#minimizing-for-production</a></li><li>How does production site is built using webpack — <a href="https://webpack.js.org/guides/production/">https://webpack.js.org/guides/production/</a></li><li>Setting up perfect devpack server — <a href="https://linguinecode.com/post/how-to-setup-webpack-dev-server-react-babel">https://linguinecode.com/post/how-to-setup-webpack-dev-server-react-babel</a></li><li>Loaders in details — <a href="https://webpack.js.org/concepts/loaders/">https://webpack.js.org/concepts/loaders/</a></li><li>Understanding babel-preset-env in details — <a href="https://blog.jakoblind.no/babel-preset-env/">https://blog.jakoblind.no/babel-preset-env/</a></li></ol><p>I hope you found this article useful. Will meet you soon with next one.</p><h3>Build React Apps with reusable components, just like Lego</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/0*PMAicpZLCKFZjKpO.png" /></figure><p><a href="https://bit.cloud/"><strong>Bit</strong></a><strong>’s open-source tool </strong>help 250,000+ devs to build apps with components.</p><p>Turn any UI, feature, or page into a <strong>reusable component</strong> — and share it across your applications. It’s easier to collaborate and build faster.</p><p><strong>→ </strong><a href="https://bit.dev/"><strong>Learn more</strong></a></p><p>Split apps into components to make app development easier, and enjoy the best experience for the workflows you want:</p><h4>→ <a href="https://blog.bitsrc.io/how-we-build-micro-front-ends-d3eeeac0acfc">Micro-Frontends</a></h4><h4>→ <a href="https://blog.bitsrc.io/how-we-build-our-design-system-15713a1f1833">Design System</a></h4><h4>→ <a href="https://bit.cloud/blog/how-to-reuse-react-components-across-your-projects-l4pz83f4">Code-Sharing and reuse</a></h4><h4>→ <a href="https://www.youtube.com/watch?v=5wxyDLXRho4&amp;t=2041s">Monorepo</a></h4><h4>Learn more:</h4><ul><li><a href="https://blog.bitsrc.io/how-we-build-micro-front-ends-d3eeeac0acfc">How We Build Micro Frontends</a></li><li><a href="https://blog.bitsrc.io/how-we-build-our-design-system-15713a1f1833">How we Build a Component Design System</a></li><li><a href="https://bit.dev/blog/how-to-reuse-react-components-across-your-projects-l4pz83f4/">How to reuse React components across your projects</a></li><li><a href="https://blog.bitsrc.io/5-ways-to-build-a-react-monorepo-a294b6c5b0ac">5 Ways to Build a React Monorepo</a></li><li><a href="https://bit.dev/blog/how-to-create-a-composable-react-app-with-bit-l7ejpfhc/">How to Create a Composable React App with Bit</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b0a5806a92" width="1" height="1" alt=""><hr><p><a href="https://medium.com/bitsrc/create-react-app-without-create-react-app-b0a5806a92">Create React App without Create React App</a> was originally published in <a href="https://medium.com/bitsrc">Bits and Pieces</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>