<?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 Chamila Ambahera on Medium]]></title>
        <description><![CDATA[Stories by Chamila Ambahera on Medium]]></description>
        <link>https://medium.com/@ambahera?source=rss-e2edbe6eb3e4------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*ZTAaNs7TvqIjz8Wcb9VhWw.jpeg</url>
            <title>Stories by Chamila Ambahera on Medium</title>
            <link>https://medium.com/@ambahera?source=rss-e2edbe6eb3e4------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 01:06:21 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ambahera/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[The Canary That Saved 11 Lives — And What It Teaches Us About Software Releases]]></title>
            <link>https://ambahera.medium.com/the-canary-that-saved-11-lives-and-what-it-teaches-us-about-software-releases-24bab303f5a1?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/24bab303f5a1</guid>
            <category><![CDATA[qa]]></category>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[software-testing]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[software-engineering]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Thu, 25 Sep 2025 15:01:49 GMT</pubDate>
            <atom:updated>2025-09-25T15:01:49.880Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Canary That Saved 11 Lives — And What It Teaches Us About Software Releases</h3><blockquote>Why a tiny bird became more trusted than decades of mining experience, and what modern development teams can learn from 1911</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XJysVV6fhAAYAWAnhcbxHw.png" /></figure><p>Deep underground in 1911, Welsh miner Thomas descended with something unusual: a small yellow canary in a wooden cage. His crew thought he’d lost his mind.</p><p>Three hours later, that bird saved eleven lives.</p><p>The canary grew restless, then stopped singing — warning that deadly carbon monoxide was seeping in. They evacuated minutes before gas levels turned lethal.</p><p>This wasn’t luck. It was the beginning of a safety revolution that would transform mining forever — and offers profound lessons for how we deploy software today.</p><h3>The Invisible Enemy That Experience Couldn’t Defeat</h3><p>Before canaries, miners relied on gut feelings and hard-earned experience. Veterans would sniff the air, watch flame patterns, trust their underground knowledge built over decades in dark tunnels.</p><p><strong>But invisible gases don’t care about experience.</strong></p><p>Carbon monoxide would build up gradually in mine shafts — no smell, no color, no warning. By the time humans felt symptoms (headaches, dizziness, confusion), it was often too late to escape. Entire crews would collapse simultaneously, with no one left to sound the alarm.</p><p>Seasoned miners died alongside newcomers when invisible threats overwhelmed human senses. Experience, it turned out, had a fatal blind spot.</p><p>The canary system forced a fundamental mindset shift: <strong>stop trusting assumptions, start trusting observable signals.</strong></p><p>A singing bird became more reliable than thirty years of mining wisdom.</p><h3>The Modern Software “Mine Shaft”</h3><p>Today’s development teams face remarkably similar invisible threats. We push code into production environments where millions of users interact with our software in ways we can’t predict or fully simulate in advance.</p><p>We rely on familiar confidence patterns:</p><ul><li>“The staging tests passed”</li><li>“We’ve done similar releases before”</li><li>“Code review looked good”</li><li>“Should be fine”</li></ul><p><strong>But production, like mine shafts, loves to surprise even experienced teams.</strong></p><p>Modern applications are complex ecosystems where subtle interactions between components, unexpected user behaviors, and real-world data patterns create the software equivalent of invisible gases. By the time we notice widespread issues — support tickets flooding in, error rates spiking, user complaints on social media — the damage is already done.</p><h3>Enter the Digital Canary</h3><p>Canary testing follows the same life-saving principle that transformed mining: <strong>expose a small group to risk first, monitor their response, make decisions based on observable signals.</strong></p><p>The process mirrors those Welsh miners’ approach:</p><ol><li><strong>Release to 5% of users first</strong> — your digital canaries</li><li><strong>Watch critical signals</strong>: error rates, response times, user engagement, support tickets</li><li><strong>Healthy signals?</strong> Expand to 25%, then 50%, then 100%</li><li><strong>Distress signals?</strong> Roll back immediately, affecting only the small group</li></ol><p>Just like those miners trusted their bird’s silence over personal judgment, modern teams must trust data over confidence in their code.</p><h3>The Psychology That Changes Everything</h3><p>When miners knew canaries would warn them of danger before it became fatal, something remarkable happened: <strong>they became willing to explore deeper, more challenging tunnels.</strong></p><p>The safety net didn’t make them reckless — it made them bold in calculated ways.</p><p>The same psychological transformation occurs with development teams using canary deployments:</p><ul><li><strong>Developers</strong> try performance improvements they’ve been nervous about, knowing failures will be caught early and contained</li><li><strong>Product managers</strong> test features they weren’t sure would work, because user feedback comes from real usage rather than hypothetical scenarios</li><li><strong>QA professionals</strong> shift focus from trying to prevent every possible failure upfront to enabling fast, safe learning in production</li></ul><p>Fear of “breaking production for everyone” stops paralyzing decision-making. Teams move from conservative, incremental changes to meaningful innovation.</p><h3>The Discipline That Makes It Work</h3><p>Here’s where many teams stumble: successful miners didn’t debate when birds went quiet — they evacuated immediately. <strong>The signal overruled confidence every time.</strong></p><p>Modern canary testing requires the same unwavering discipline. When 5% of users show distress signals, you roll back regardless of how confident you feel about the code. When metrics indicate problems, you trust the data over explanations about why the problems “shouldn’t” be happening.</p><p>The system only works when teams commit to following signals instead of rationalizing them away.</p><p>This is harder than it sounds. Engineers naturally want to investigate, debug, and fix issues rather than retreat. Product managers worry about delaying features. Business stakeholders question the need for rollbacks when “most users seem fine.”</p><p>But the canary’s power lies in its early warning capacity. Ignore the warning, and you lose the entire advantage.</p><h3>What Those Miners Knew That We’re Still Learning</h3><p>The most successful mining operations didn’t see canaries as just safety tools — <strong>they saw them as exploration enablers.</strong></p><p>The birds didn’t prevent all risks. They made risks manageable, observable, and recoverable.</p><p>With canaries providing reliable early warning systems, mining operations expanded into previously uncharted territories. They could explore confidently because they had safety nets that would alert them to danger before it became catastrophic.</p><p>This is the real lesson for modern development teams: canary testing isn’t about avoiding failure. <strong>It’s about failing safely, learning quickly, and building systems that get stronger through small, controlled experiments rather than big, risky bets.</strong></p><p>Teams using canary deployments consistently report catching issues that would have been invisible in traditional testing:</p><ul><li>Subtle performance regressions affecting specific user segments</li><li>Feature interactions that only emerge under real-world usage patterns</li><li>Edge cases that no amount of staging environment testing would reveal</li></ul><p>More importantly, they report increased confidence in their releases and improved collaboration across departments.</p><h3>The Trust That Enables Innovation</h3><p>Those Welsh miners made a profound leap of faith: they chose to trust a canary’s behavior over their own experience and instincts. This wasn’t about abandoning expertise — it was about <strong>recognizing the limits of human perception and creating better feedback loops.</strong></p><p>Modern software teams need the same humility and wisdom. Our staging environments, no matter how sophisticated, are elaborate simulations of production. Our test cases, no matter how comprehensive, are educated guesses about user behavior. Our architectural reviews, no matter how thorough, can’t predict every interaction in complex distributed systems.</p><p>Canary testing acknowledges these limitations while providing a path forward: <strong>use real users and real production environments as your ultimate validation, but do it safely and gradually.</strong></p><h3>Getting Started: From Concept to Practice</h3><p>The beauty of canary testing is that you don’t need sophisticated infrastructure to begin. Start with one upcoming feature release:</p><ol><li><strong>Define clear success metrics</strong> before deploying (error rates, response times, user engagement)</li><li><strong>Set specific rollback criteria</strong> that everyone agrees on upfront</li><li><strong>Create shared dashboards</strong> so all team members watch the same signals</li><li><strong>Start small</strong>: 5% of users, basic monitoring, clear decision points</li></ol><p>The goal isn’t perfection — it’s <strong>fast, safe learning</strong> with real user feedback.</p><h3>The Question Every Team Should Ask</h3><p>If Welsh miners in 1911 could trust a canary’s silence over decades of underground experience, what signals should modern development teams trust over their confidence in code quality?</p><p>The answer reveals everything about how we approach risk, learning, and innovation in software development.</p><p>Your users might not know they’re serving as your canaries, but they’re helping you build better, safer software — one small, monitored release at a time. The question is whether you’re listening to what they’re telling you.</p><p><em>What early warning signals does your team rely on? Have you experienced the psychological shift that comes from knowing failures will be small and contained? Share your thoughts in the comments below.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=24bab303f5a1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 4]]></title>
            <link>https://ambahera.medium.com/learn-istqb-foundation-level-7x-faster-smarter-not-harder-part-4-887be42c4340?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/887be42c4340</guid>
            <category><![CDATA[istqb-foundation-level]]></category>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[qa]]></category>
            <category><![CDATA[qa-testing-training]]></category>
            <category><![CDATA[software-testing]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Tue, 19 Aug 2025 08:02:05 GMT</pubDate>
            <atom:updated>2025-08-19T08:02:05.499Z</atom:updated>
            <content:encoded><![CDATA[<h3>Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 4</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/770/0*FagBl_3vvJTTweyi.png" /></figure><p>Got an interview coming up?<br>Need a quick refresher on the QA process?</p><p>Or maybe you want to master the ISTQB foundation level in record time?<br>Good News — You have just found the perfect guide!!</p><p>Let’s be real — tired of watching boring technical videos on ISTQB?<br>But what if learning <strong>Software Quality Assurance (QA)</strong> felt more like watching a great story unfold?</p><p>That’s exactly what the <strong>TestTactix Media Team</strong> has done! 🚀 They’ve created a <strong>YouTube series</strong> that teaches QA <strong>through storytelling</strong> — no dull lectures, just <strong>real-world examples, simplified concepts, and engaging narratives</strong> that actually make learning fun.</p><p>If you’re looking for an exciting way to master QA, this is <strong>definitely worth your time!</strong></p><p>If you haven’t checked Part 1 of this series, please check below.</p><p><a href="https://ambahera.medium.com/learn-istqb-foundation-level-7x-faster-smarter-not-harder-part-1-7be3cf592a10">Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 1</a></p><h3><strong>Black-Box Test Techniques Made Simple! 🎯</strong></h3><p>Join Sara &amp; Bob as they demystify the most essential black-box test design techniques — Equivalence Partitioning, Boundary Value Analysis, Decision Tables, and State Transition Testing. Whether you’re prepping for ISTQB Foundation or refining your QA skills, this episode gives you the clarity you need to apply these techniques with confidence.</p><p><strong>Here’s what you’ll master in this episode:</strong><br> 1️⃣ Equivalence Partitioning (EP) — Reduce test cases while keeping coverage strong.<br> 2️⃣ Boundary Value Analysis (BVA) — Catch edge-case bugs before they bite.<br> 3️⃣ Decision Tables — Handle complex business rules without missing scenarios.<br> 4️⃣ State Transition Testing — Test systems that change behavior over time.</p><p><strong>🔑 What You’ll Learn:</strong><br> ✅ When and why to use each black-box technique.<br> ✅ Real-world examples for EP, BVA, Decision Tables, and State Transitions.<br> ✅ How to avoid common pitfalls when applying these methods.<br> ✅ How these techniques help you focus on <em>what</em> the system does, not how it works.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FA6OAJejEF84%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DA6OAJejEF84&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FA6OAJejEF84%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/fe23859b9d08c658f852276964764035/href">https://medium.com/media/fe23859b9d08c658f852276964764035/href</a></iframe><h3>Crack open the code and test with precision! 💻</h3><p>In this episode, Sara teaches Bob (and you!) the power of White-Box Testing Techniques — your secret weapon for bulletproof code quality.</p><p>Here’s what we’ll cover:<br>1️⃣ Statement Testing — Execute every line of code (but is that enough?)<br>2️⃣ Branch Testing — Conquer every if/else path (the real key to coverage!)<br>3️⃣ Code Coverage Explained — Why 100% statement ≠ 100% safety<br>4️⃣ Real Code Examples — See exactly how to test conditional logic<br>5️⃣ When to Use White-Box — The perfect complement to black-box testing</p><p>🔑 Key Takeaways You’ll Get:<br>✅ The critical difference between statement and branch coverage<br>✅ How to find hidden defects that black-box tests miss<br>✅ Why branch coverage “subsumes” statement coverage<br>✅ Practical examples with Python-style code snippets<br>✅ Pro tips for balancing white-box and black-box approaches</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F6jKHBZeWP54%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D6jKHBZeWP54&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F6jKHBZeWP54%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/92cd5cb06e1ac22ac665305f24535863/href">https://medium.com/media/92cd5cb06e1ac22ac665305f24535863/href</a></iframe><h3>🔍 Experience-Based Testing: Why Manual QA Skills Still Matter</h3><p>In an era of automation, manual testing remains indispensable. Join Sara and Bob as they break down proven experience-based techniques that leverage human expertise to find critical bugs.</p><p>📌 Evidence-Based Techniques Covered:<br>1️⃣ Error Guessing — Using historical knowledge to predict defect-prone areas <br>2️⃣ Exploratory Testing — Unstructured testing that adapts to discoveries <br>3️⃣ Checklist-Based Testing — Balanced approach for coverage without rigidity<br>4️⃣ Collaboration Methods — User stories and acceptance criteria for defect prevention<br>5️⃣ ATDD (Acceptance Test-Driven Development) — Defining tests before implementation</p><p>🔬 Documented Benefits of Manual QA:<br>✅ Identifies UI/UX issues automation often misses (ISTQB Foundation Level Syllabus)<br>✅ Essential for exploratory testing of complex systems (IEEE Software Journal)<br>✅ Critical in agile environments with evolving requirements (Agile Testing: A Practical Guide)</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F1QFZHZ-L_qk%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D1QFZHZ-L_qk&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F1QFZHZ-L_qk%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/33e4c660d1f81c021983719ef1afb015/href">https://medium.com/media/33e4c660d1f81c021983719ef1afb015/href</a></iframe><h3>Test Planning Secrets: From Chaos to Confidence (Real-Life QA Lessons!)</h3><p>Struggle with messy test cycles? ☕ Join Sara and Bob in their café chat to learn test planning strategies that work for both software and life — proving even “go-with-the-flow” testers can master organization!</p><p>📌 In This Video:<br>✅ The “Wedding Planning” Approach — Objectives, risks &amp; stakeholder alignment<br>✅ Release vs Iteration Planning — Road trips vs snack packing analogies<br>✅ Estimation Hacks — Why Planning Poker beats wild guesses<br>✅ Entry/Exit Criteria — The “toothpick test” for software quality<br>✅ Test Pyramid Explained — Why foundations matter more than UI glitter<br>✅ Prioritization Secrets — Clean your apartment (or tests) the smart way</p><p>🔑 Key Takeaways:<br>🚀 Transform reactive testing into strategic QA<br>💡 Use real-world analogies (cooking, road trips) to explain complex concepts<br>📊 Balance speed and coverage with the test pyramid<br>🛡️ Anticipate risks like a pro (no more “undercooked noodles” disasters!)</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FSYMAfv8Qamg%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DSYMAfv8Qamg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FSYMAfv8Qamg%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/cbe0dcbff4a8b85b2e4ba07a071d7b01/href">https://medium.com/media/cbe0dcbff4a8b85b2e4ba07a071d7b01/href</a></iframe><h3>Risk Management in Testing: How to Prevent Software Disasters</h3><p>Join Sara and Bob in their cozy café as they break down risk management strategies that every tester needs to know — from identifying project risks to preventing product failures that could cost millions!</p><p>📌 In This Episode:<br>✅ The 2 Components of Every Risk (Likelihood + Impact)<br>✅ Project Risks vs Product Risks — Which Keeps CEOs Awake?<br>✅ Real-World Analogies (Why “Crossing Blindly” = Bad Testing)<br>✅ Risk Analysis Framework — How Netflix Prioritizes Threats<br>✅ 5 Pro Mitigation Strategies (The “Fire Extinguisher” Approach)<br>✅ When Good Testing Prevents Lawsuits (True Case Study)</p><p>🔑 Key Takeaways:<br>🚀 Transform from reactive to proactive testing<br>💡 Learn how risk-based testing saves 40% effort (ISTQB Data)<br>📊 Master the risk assessment matrix used by FAANG companies<br>🛡️ Discover why some risks are worth accepting</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fps87cefOOoE%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dps87cefOOoE&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fps87cefOOoE%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/1aed4a845571320c018426d0ebcb2415/href">https://medium.com/media/1aed4a845571320c018426d0ebcb2415/href</a></iframe><h3>Test Monitoring, Control &amp; Completion Explained with Real-World QA Analogies</h3><p>Unlock the key to mastering <strong>Test Monitoring, Test Control, and Test Completion</strong> in this engaging episode with Sara and Bob! 🚀 Whether you’re preparing for ISTQB certification, gearing up for interviews, or sharpening your QA fundamentals, this one’s packed with practical insights.</p><p>Here’s what you’ll master in this episode:<br> 1️⃣ <strong>Test Monitoring</strong> — Tracking progress, quality, defects, and risks in real time.<br> 2️⃣ <strong>Test Control</strong> — Making course corrections to keep testing on track.<br> 3️⃣ <strong>Key Test Metrics</strong> — The numbers that matter most for quality and decision-making.<br> 4️⃣ <strong>Test Progress &amp; Completion Reports</strong> — Communicating results effectively to stakeholders.</p><p>🔑 What You’ll Learn:<br> ✅ The clear difference between <strong>monitoring</strong> and <strong>control</strong> (with easy analogies).<br> ✅ How to use test metrics to make smarter, faster decisions.<br> ✅ Tips for creating reports that actually get read (and acted on).<br> ✅ How to tailor communication for managers, developers, and clients.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FjPsNBvd27bs%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DjPsNBvd27bs&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FjPsNBvd27bs%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/6f34aab43b1353fe702d89029b444553/href">https://medium.com/media/6f34aab43b1353fe702d89029b444553/href</a></iframe><h3>Confused about <strong>Configuration Management</strong> and <strong>Defect Management</strong> in Software Testing?</h3><p>In this episode, Sara and Bob break down these critical QA processes with <strong>simple analogies</strong> and <strong>practical insights</strong> you can apply right away. Perfect for ISTQB prep, QA interviews, or anyone starting their testing career.</p><p>Here’s what you’ll master in this episode:<br> 1️⃣ <strong>Configuration Management</strong> — Keeping your test assets, environments, and documents consistent and under control.<br> 2️⃣ <strong>Baselines &amp; Version Control</strong> — Why they matter for smooth test execution.<br> 3️⃣ <strong>Defect Management</strong> — More than just finding bugs — it’s about tracking and improving.<br> 4️⃣ <strong>Defect Lifecycle</strong> — From detection to closure, step by step.</p><p>🔑 What You’ll Learn:<br> ✅ How configuration management ensures reliable and repeatable testing.<br> ✅ The critical role of baselines and version history in QA.<br> ✅ How to write clear, complete, and actionable defect reports.<br> ✅ Real-world defect scenarios explained with humor and clarity.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FE5g1G3ZKHF0%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DE5g1G3ZKHF0&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FE5g1G3ZKHF0%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/e86ec38f485d9a0fa2bcb13a5e879aaf/href">https://medium.com/media/e86ec38f485d9a0fa2bcb13a5e879aaf/href</a></iframe><h3><strong>End of a QA Journey… Or Just the Beginning? | Test Automation &amp; QA Tools Explained</strong></h3><p>In this <strong>final episode</strong> of our QA learning journey, Sara guides Bob through the exciting (and sometimes misunderstood) world of <strong>Test Automation</strong> and <strong>QA tools</strong>. 🤖💻<strong><br>Here’s what you’ll master in this episode:</strong><br> 1️⃣ What test automation is <em>really</em> about (beyond the hype)<br> 2️⃣ Real benefits and risks of automation tools<br> 3️⃣ Why automation can’t fully replace manual testing<br> 4️⃣ How to avoid common tool selection pitfalls<br> 5️⃣ Why human intuition and strategy still matter in QA</p><p><strong>🔑 What You’ll Learn:</strong><br> ✅ The role of test tools in the software lifecycle<br> ✅ Why automation is powerful — but not magical<br> ✅ Common mistakes teams make when choosing tools<br> ✅ How automation complements (not replaces) manual testing<br> ✅ Sara’s final advice for every aspiring tester 💡</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FlLVJNqQUQN4%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DlLVJNqQUQN4&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FlLVJNqQUQN4%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/e8f071ee878bd070dd47d3118c6a9630/href">https://medium.com/media/e8f071ee878bd070dd47d3118c6a9630/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=887be42c4340" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Good Test Automation Doesn’t Start with Code]]></title>
            <link>https://ambahera.medium.com/good-test-automation-doesnt-start-with-code-c5db5fb05f4d?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/c5db5fb05f4d</guid>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[test-automation]]></category>
            <category><![CDATA[test-automation-framework]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[software-testing]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Tue, 13 May 2025 10:01:40 GMT</pubDate>
            <atom:updated>2025-05-13T10:01:40.820Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Before diving into scripts and frameworks, I learned that the most impactful test automation starts with asking the right questions, not writing code.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YGAzx-wZwvmuhxQlcxvcXQ.png" /></figure><p>When I first got into test automation, I thought the goal was obvious:</p><p>Write clean scripts. Learn all the latest tools. Build slick CI/CD pipelines.</p><p>So that’s exactly what I did.</p><p>I buried myself in frameworks, automated everything I could, and proudly built dashboards to show off the test runs.</p><p>But after a while, I noticed something… odd.</p><p>🚫 Despite all the code and tools, product quality didn’t really improve.<br> ✅ And the automation wins that <em>did</em> matter? They had nothing to do with how clever my code was.</p><p>They started with something much simpler:</p><p><strong>Conversations.</strong></p><h3>These days, I don’t jump into code. I start with clarity.</h3><p>Before I write a single line, I try to pause and think. Here’s what I focus on instead:</p><h3>🔍 Start by asking: What actually needs testing?</h3><p>Not everything needs to be automated. Some things are better left manual, or even ignored.</p><p>Now, I ask myself:</p><blockquote><em>“If this breaks, who will notice? And how soon?”</em></blockquote><p>This one question helps me focus on the flows that actually matter — the ones tied to real users and real value.<br> No more wasting time automating obscure edge cases no one cares about.</p><h3>🧠 Be clear on the why</h3><p>What are we even trying to achieve?</p><p>Regression coverage? Smoke checks on each deploy? Monitoring the core user journey in production?</p><p>Each of these goals needs a different approach.<br> Knowing this upfront has saved me from so much trial-and-error (and a lot of rework later).</p><h3>🤝 Don’t do it alone</h3><p>One of the biggest shifts for me?</p><p>I stopped treating automation like my personal solo project.</p><p>Now I pull in devs, QAs, product folks, even the support team when I can.<br> They all bring perspectives I would’ve missed — and that makes the automation better.</p><p>When everyone’s on board, the tests are more accurate, useful, and sustainable.</p><h3>📐 Design it like you mean it</h3><p>I used to jump straight into code. Now? I take a breath and design first.</p><p>Just like we don’t build features without a design doc, I don’t build tests without a plan.</p><p>I sketch out how the test code will work:</p><ul><li>How do we keep it clean and easy to maintain?</li><li>Can we keep test data independent?</li><li>Will results be clear and visible?</li></ul><p>This upfront thinking makes everything smoother later on.</p><p>So now, when someone says, “Let’s start automating,”</p><p><em>I don’t open IntelliJ.</em></p><p><em>I open Notepad.</em></p><p><em>And we start asking questions.</em></p><p>Because in the end:</p><blockquote><strong><em>Good test automation doesn’t start with code — it starts with clarity.</em></strong></blockquote><p><strong>I’m curious — how do you approach test automation planning?</strong><br> Drop your thoughts below 👇 Would love to learn from your experience too.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c5db5fb05f4d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 3]]></title>
            <link>https://ambahera.medium.com/learn-istqb-foundation-level-7x-faster-smarter-not-harder-part-3-7537cefb78f0?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/7537cefb78f0</guid>
            <category><![CDATA[qa]]></category>
            <category><![CDATA[software-testing]]></category>
            <category><![CDATA[istqb]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[qa-testing]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Tue, 08 Apr 2025 10:01:57 GMT</pubDate>
            <atom:updated>2025-04-08T10:01:57.754Z</atom:updated>
            <content:encoded><![CDATA[<h3>Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 3</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S5dhiCCYecKtXUvsWd6BHQ.png" /></figure><p>Got an interview coming up?<br>Need a quick refresher on the QA process?</p><p>Or maybe you want to master the ISTQB foundation level in record time? <br>Good News — You have just found the perfect guide!!</p><p>Let’s be real — tired of watching boring technical videos on ISTQB?<br>But what if learning <strong>Software Quality Assurance (QA)</strong> felt more like watching a great story unfold?</p><p>That’s exactly what the <strong>TestTactix Media Team</strong> has done! 🚀 They’ve created a <strong>YouTube series</strong> that teaches QA <strong>through storytelling</strong> — no dull lectures, just <strong>real-world examples, simplified concepts, and engaging narratives</strong> that actually make learning fun.</p><p>If you’re looking for an exciting way to master QA, this is <strong>definitely worth your time!</strong></p><p>If you haven’t checked Part 1 of this series please check below.</p><p><a href="https://ambahera.medium.com/learn-istqb-foundation-level-7x-faster-smarter-not-harder-part-1-7be3cf592a10">Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 1</a></p><h3>Static Testing Basics</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FqbV9DWFmtyI%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DqbV9DWFmtyI&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FqbV9DWFmtyI%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/95025f404cb8a53f0c1e5fa83e69cc2e/href">https://medium.com/media/95025f404cb8a53f0c1e5fa83e69cc2e/href</a></iframe><p>📌 You’ll Learn:</p><p>1️⃣ What is Static Testing? — Detect issues before execution with code reviews, static analysis, and documentation checks</p><p>2️⃣ Review Techniques — Strengthen requirements, test cases, and design with structured walkthroughs &amp; inspections</p><p>3️⃣ Static Analysis Tools — Use automation to detect security vulnerabilities, coding standards violations, and performance issues</p><p>4️⃣ Collaboration in Static Testing — Improve Agile, DevOps, and CI/CD processes by aligning QA, developers, and business teams early</p><p>5️⃣ Cost Savings &amp; Efficiency — Reduce software development costs by catching defects early and minimizing late-stage bug fixes</p><h3>Differences between Static Testing and Dynamic Testing</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FRZnAJyIh4DY%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DRZnAJyIh4DY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FRZnAJyIh4DY%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/286e23a263fef638765cff36901b94c9/href">https://medium.com/media/286e23a263fef638765cff36901b94c9/href</a></iframe><p>📌 You’ll Learn:</p><p>✅ Static Testing:</p><p>Review requirements, code, and designs without execution</p><p>Catch defects early (ambiguities, security flaws, unreachable code)</p><p>Tools: Manual reviews, static analysis tools, pair programming</p><p>✅ Dynamic Testing:</p><p>Validate software behavior during runtime</p><p>Test performance, edge cases, and real-world scenarios</p><p>Tools: Unit testing frameworks (JUnit, pytest), load testing tools</p><h3>Reviews &amp; Feedback</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FdGrCFA1oz_M%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DdGrCFA1oz_M&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FdGrCFA1oz_M%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4ee60052a12ebe2417b17f522683484e/href">https://medium.com/media/4ee60052a12ebe2417b17f522683484e/href</a></iframe><p>🔑 What You’ll Learn:</p><p>✅ How early feedback saves time and avoids disasters.</p><p>✅ Real-world examples of review types and when to use them.</p><p>✅ Tips to keep reviews focused, productive, and conflict-free.</p><p>✅ The secret to fostering a culture of continuous improvement.</p><p>Let me know what you think about this new way of learning in the comment section.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7537cefb78f0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 2]]></title>
            <link>https://ambahera.medium.com/learn-istqb-foundation-level-7x-faster-smarter-not-harder-part-2-2ca205bb0eea?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/2ca205bb0eea</guid>
            <category><![CDATA[qa-testing]]></category>
            <category><![CDATA[istqb]]></category>
            <category><![CDATA[software-testing]]></category>
            <category><![CDATA[istqb-foundation-level]]></category>
            <category><![CDATA[qa]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Mon, 31 Mar 2025 10:02:11 GMT</pubDate>
            <atom:updated>2025-03-31T10:02:11.097Z</atom:updated>
            <content:encoded><![CDATA[<h3>Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 2</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S5dhiCCYecKtXUvsWd6BHQ.png" /></figure><p>Got an interview coming up?<br>Need a quick refresher on the QA process?</p><p>Or maybe you want to master the ISTQB foundation level in record time?</p><p>Good News — You have just found the perfect guide!!</p><p>Let’s be real — tired of watching boring technical videos on ISTQB?<br>But what if learning <strong>Software Quality Assurance (QA)</strong> felt more like watching a great story unfold?</p><p>That’s exactly what the <strong>TestTactix Media Team</strong> has done! 🚀 They’ve created a <strong>YouTube series</strong> that teaches QA <strong>through storytelling</strong> — no dull lectures, just <strong>real-world examples, simplified concepts, and engaging narratives</strong> that actually make learning fun.</p><p>If you’re looking for an exciting way to master QA, this is <strong>definitely worth your time!</strong></p><p>If you haven’t checked Part 1 of this series please check below.</p><p><a href="https://ambahera.medium.com/learn-istqb-foundation-level-7x-faster-smarter-not-harder-part-1-7be3cf592a10">Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 1</a></p><h3>Testing in the SDLC</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FaFTPFkUGF8E%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DaFTPFkUGF8E&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FaFTPFkUGF8E%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/0da31e1297ee97b2c55b445826b5dfe8/href">https://medium.com/media/0da31e1297ee97b2c55b445826b5dfe8/href</a></iframe><p><strong>🔍 What You’ll Learn:</strong></p><ul><li>What is SDLC? A crash course in Waterfall, V-Model, Agile, and Iterative Models.</li><li>The pros and cons of each SDLC model for testing.</li><li>How Agile integrates continuous testing into every phase.</li><li>Why testing isn’t just about finding bugs — it’s about building confidence and guiding teams.</li><li>The role of regression testing in Agile for catching issues early.</li></ul><h3>TDD, ATDD, BDD, and DevOps Practices Explained</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Frgzp0T4PR-E%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Drgzp0T4PR-E&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Frgzp0T4PR-E%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/a9936019c151f82a5c858f45d9c74264/href">https://medium.com/media/a9936019c151f82a5c858f45d9c74264/href</a></iframe><p><strong>💡 You’ll Learn</strong></p><ul><li>How TDD, ATDD, and BDD improve collaboration</li><li>The role of DevOps in integrating development, testing, and operations</li><li>Benefits and challenges of automated testing in a CI/CD pipeline</li><li>Why testing is more than finding bugs — it’s about building quality from the start</li></ul><h3>Shift-Left Testing</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FRJMrMDW7btY%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DRJMrMDW7btY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FRJMrMDW7btY%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/5fc7a44027bdd2c029c2733bbb517277/href">https://medium.com/media/5fc7a44027bdd2c029c2733bbb517277/href</a></iframe><p>💡 What You’ll Learn:</p><ul><li>Why Shift-Left Testing is a game-changer — Reduce late-stage firefighting by 50%!</li><li>Actionable strategies to embed testing into requirements, design, and coding phases.</li><li>Boost collaboration between teams with shared ownership of quality.</li><li>Leverage CI/CD pipelines for instant feedback and automated testing.</li><li>Run retrospectives that drive continuous improvement (and team bonding!).</li></ul><h3>Test Levels</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FKHKWU2npXus%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DKHKWU2npXus&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FKHKWU2npXus%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/a5b2d88df62cf885215adce4bb8e5152/href">https://medium.com/media/a5b2d88df62cf885215adce4bb8e5152/href</a></iframe><p>📌You’ll Learn:</p><ul><li>Unit Testing — Validate individual functions/methods with tools like JUnit, pytest, or NUnit</li><li>Component Testing — Isolate and verify entire modules or features</li><li>Component Integration Testing — Master strategies (top-down, bottom-up, big-bang) for smooth component interactions</li><li>System Testing — Stress-test end-to-end functionality &amp; non-functional requirements (performance, security)</li><li>System Integration Testing — Ensure compatibility with APIs, third-party services, and legacy systems</li><li>Acceptance Testing — Align with stakeholders via UAT, alpha/beta testing, and operational checks</li></ul><h3>Test Types</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FoXga8gcaJx8%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DoXga8gcaJx8&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FoXga8gcaJx8%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/dd8d6e15f2d89317ee060dd2ec991157/href">https://medium.com/media/dd8d6e15f2d89317ee060dd2ec991157/href</a></iframe><p>📌 What You’ll Learn:</p><ul><li>Functional Testing — Focus on the ‘what’ of the system (e.g., login functionality)</li><li>Non-Functional Testing — Evaluate the ‘how’ (e.g., performance, usability)</li><li>Black-Box Testing — Test external behaviour without peeking into the code</li><li>White-Box Testing — Dive into internal structures like code and workflows</li><li>Confirmation &amp; Regression Testing — Validate fixes and prevent new bugs</li></ul><h3>Maintenance Testing</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F21ir3iuhHfc%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D21ir3iuhHfc&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F21ir3iuhHfc%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/b45ab36b03d21d2545a95088047df536/href">https://medium.com/media/b45ab36b03d21d2545a95088047df536/href</a></iframe><p>📌 What You’ll Learn:</p><ul><li>Types of Maintenance: Dive into corrective, adaptive, and performance improvement testing</li><li>Maintenance Triggers: From urgent hotfixes to planned upgrades and system retirements</li><li>Impact Analysis: Learn to prioritize testing efforts based on risk assessment</li><li>Regression Prevention: Why testing changes in production is crucial for system stability</li><li>Data Integrity: Explore the importance of migration, archiving, and restoration testing</li></ul><p>Let me know what you think about this new way of learning in the comment section.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2ca205bb0eea" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 1]]></title>
            <link>https://ambahera.medium.com/learn-istqb-foundation-level-7x-faster-smarter-not-harder-part-1-7be3cf592a10?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/7be3cf592a10</guid>
            <category><![CDATA[testing]]></category>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[test-automation]]></category>
            <category><![CDATA[qa-testing]]></category>
            <category><![CDATA[qa]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Tue, 11 Feb 2025 07:01:45 GMT</pubDate>
            <atom:updated>2025-02-11T07:01:45.778Z</atom:updated>
            <content:encoded><![CDATA[<h3>Learn ISTQB Foundation Level 7X Faster — Smarter, Not Harder! — Part 1</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Bz4kiE0FL68_MHbt9Z8HSQ.png" /></figure><p>Let’s be real — tired of watching boring technical videos on ISTQB?<br>But what if learning <strong>Software Quality Assurance (QA)</strong> felt more like watching a great story unfold?</p><p>That’s exactly what the <strong>TestTactix Media Team</strong> has done! 🚀 They’ve created a <strong>YouTube series</strong> that teaches QA <strong>through storytelling</strong> — no dull lectures, just <strong>real-world examples, simplified concepts, and engaging narratives</strong> that actually make learning fun.</p><p>If you’re looking for an exciting way to master QA, this is <strong>definitely worth your time!</strong></p><p>Why 7x faster? Why not 20x or 10x? Find the answer at the end.</p><p>This set of videos will cover “Fundamentals of Testing” in the ISTQB Foundation syllabus.</p><h4>Software Testing 101: Lessons from Bob’s First Day as a QA Engineer</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FtcAygQCb-Hs%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DtcAygQCb-Hs&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FtcAygQCb-Hs%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/942b820e94cf0dc041c316d2245fa330/href">https://medium.com/media/942b820e94cf0dc041c316d2245fa330/href</a></iframe><p>💡 What You’ll Learn in This Video:</p><ul><li>Why software testing is critical for quality and user satisfaction</li><li>Common misconceptions about testing and what it truly involves</li><li>The difference between verification and validation in testing</li><li>Essential tips for new software testers to succeed</li></ul><h4>Why We Test: Bob Learns Testing Objectives for CodeQuest Airlines!</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fzya8RFkdB3o%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dzya8RFkdB3o&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fzya8RFkdB3o%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/48e119e895516aca5475a0b29f3318d7/href">https://medium.com/media/48e119e895516aca5475a0b29f3318d7/href</a></iframe><p>✍️ What You’ll Learn:</p><ul><li>Why testing is about more than just finding bugs.</li><li>How test objectives depend on the level (unit, system) and methodology (Agile, Waterfall).</li><li>The role of testing in building stakeholder confidence and verifying compliance.</li><li>Practical tips on balancing objectives with business deadlines and priorities.</li></ul><h4>Bob Found Something Weird | Testing vs. Debugging Explained</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FCgOyOlZFVJg%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DCgOyOlZFVJg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FCgOyOlZFVJg%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/892ab61929d68a336ea9a8c908cc27d0/href">https://medium.com/media/892ab61929d68a336ea9a8c908cc27d0/href</a></iframe><p>📌 What You’ll Learn:</p><ul><li>The difference between testing and debugging: why testers identify failures and developers fix them.</li><li>The three-step debugging process: reproducing the issue, diagnosing the defect, and resolving it.</li><li>Why static testing saves time by identifying defects early without triggering failures.</li><li>How confirmation testing and regression testing ensure fixes don’t break other parts of the system.</li></ul><h4>Testing is the Pre-Flight Check Your Software Needs</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FKuCm5KkoVxM%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DKuCm5KkoVxM&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FKuCm5KkoVxM%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/fc0d1b477461754a94d079579c00e2d3/href">https://medium.com/media/fc0d1b477461754a94d079579c00e2d3/href</a></iframe><p>📌 What You’ll Learn:</p><ul><li>Why Testing is Essential: Discover how testing is like a pre-flight check, ensuring software is ready for launch.</li><li>Contributions of Testing: Learn how testing improves quality, ensures user satisfaction, and evaluates readiness throughout the lifecycle.</li><li>Testing vs. QA: Understand the difference between process-oriented QA and product-focused testing.</li><li>Errors, Defects, and Failures: See how testing identifies failures, traces root causes, and prepares for external risks.</li><li>Building Confidence: Explore how testing creates trust for stakeholders and ensures smooth operations.</li></ul><h4>7 Testing Secrets Every QA Engineer Should Know (7 Testing Principles)</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fk_3IUsnkpqY%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dk_3IUsnkpqY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fk_3IUsnkpqY%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/2575a74b6b5d75d14f016f13afbf9ba0/href">https://medium.com/media/2575a74b6b5d75d14f016f13afbf9ba0/href</a></iframe><p>📌 What You’ll Learn:</p><ul><li>The 7 Testing Principles are explained with real-life analogies and humour.</li><li>Why “testing shows the presence, not the absence, of defects” and what it means for your QA process.</li><li>The secrets of the Pesticide Paradox and how to keep your tests fresh and effective.</li><li>How context, clustering, and early testing save time, money, and effort.</li><li>Why even a bug-free product can fail without meeting user needs.</li></ul><h4>Testing Frameworks Unveiled: The Wild Journey Behind Testing Frameworks</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FaMa3X9RKlaQ%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DaMa3X9RKlaQ&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FaMa3X9RKlaQ%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/16cc48568aa8fcbd5b1e2d61b55f2bfc/href">https://medium.com/media/16cc48568aa8fcbd5b1e2d61b55f2bfc/href</a></iframe><p>📌 What You’ll Learn:</p><ul><li>The Framework of Testing: Explore the stages of test planning, monitoring, analysis, design, implementation, execution, and completion.</li><li>Testing Analogies: Learn how testing compares to fishing, treasure hunting, and stargazing — engaging metaphors for an otherwise technical process.</li><li>Why Testing Matters: Discover how preparation and execution lead to successful bug hunting and robust software.</li><li>The Human Side of QA: See how Bob’s personal story adds a touching layer to the world of quality assurance.</li></ul><h4>Essential Skills and Good Practices in Testing</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FwhDELFvIXCI%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DwhDELFvIXCI&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FwhDELFvIXCI%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/d84b8cccce10e42e9154de37973b8314/href">https://medium.com/media/d84b8cccce10e42e9154de37973b8314/href</a></iframe><p>✨ What You’ll Learn:</p><ul><li>How contextual factors like stakeholders, team skills, business domain, and tools shape testing strategies.</li><li>What is testware, and why is configuration management key to staying organized?</li><li>The importance of traceability for risk assessment, coverage, and progress tracking.</li><li>How do the roles in testing (test management vs. hands-on testing) adapt to Agile environments.</li></ul><p>If you come to the end you might have a doubt about how we calculated 7x.</p><p>According to the standard ISTQB syllabus (v4, page 4), you will need to spend 180 minutes learning the Fundamentals of Testing. However, you can learn the same content using TestTactix Foundation Level videos in just 25 minutes.</p><p>Let me know what you think about this new way of learning in the comment section.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7be3cf592a10" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[10 years old Kids Vs. AI]]></title>
            <link>https://ambahera.medium.com/10-years-old-kids-vs-ai-e683e30b5950?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/e683e30b5950</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[mathamatics]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ai-model]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Thu, 06 Feb 2025 08:01:42 GMT</pubDate>
            <atom:updated>2025-02-06T08:01:42.257Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OCStmWvSn4Ls0deAS18R0Q.png" /></figure><p>With the boom of DeepSeek, people started to discuss the capabilities of artificial intelligence again. I was curious to find out if AI models could solve math problems suitable for 10-year-old kids.</p><p>I decided to use the most famous free models available today.</p><p>This question is from the Kangaroo Math competition for 9 years and 10 years old school children.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/677/1*rhi2nOwq6YIUTNzB3DeSJg.png" /></figure><p>Before diving into the rest of the article, I invite you to take a moment to tackle this question on your own. It could be a fun challenge! 😊</p><h3>ChatGPT (GPT-4o model)</h3><p>Following is the output of ChatPGT.</p><p><em>To solve this problem, let us define the variables and equations based on the given conditions:</em></p><h4><em>Variables:</em></h4><ul><li><em>TT: Sum of the numbers covered with triangles.</em></li><li><em>SS: Sum of the numbers covered with squares.</em></li><li><em>CC: The number covered with the circle.</em></li></ul><h4><em>Conditions:</em></h4><ol><li><em>T=ST = S (The sum of numbers covered with triangles equals the sum of numbers covered with squares).</em></li><li><em>C=14(T+S)C = \frac{1}{4}(T + S) (The number covered with the circle is one-quarter of the sum of TT and SS).</em></li><li><em>The total sum of numbers from 1 to 8 is: 1+2+3+4+5+6+7+8=361 + 2 + 3 + 4 + 5 + 6 + 7 + 8 = 36 Therefore, the sum of T+S+CT + S + C must be less than or equal to 36.</em></li></ol><h4><em>Goal:</em></h4><p><em>We need to find T+CT + C, the sum of numbers covered with triangles and the circle.</em></p><h4><em>Solving:</em></h4><p><em>Substitute T=ST = S into the second condition:</em></p><p><em>C=14(T+T)=14(2T)=T2.C = \frac{1}{4}(T + T) = \frac{1}{4}(2T) = \frac{T}{2}.</em></p><p><em>Thus:</em></p><p><em>T+S+C=T+T+T2=36.T + S + C = T + T + \frac{T}{2} = 36.</em></p><p><em>Simplify:</em></p><p><em>2T+T2=36.2T + \frac{T}{2} = 36.</em></p><p><em>Multiply through by 2 to eliminate the fraction:</em></p><p><em>4T+T=72 ⟹ 5T=72.4T + T = 72 \implies 5T = 72.</em></p><p><em>Solve for TT:</em></p><p><em>T=725=14.4.T = \frac{72}{5} = 14.4.</em></p><p><em>Check if integer values are consistent or reattempt solving.</em></p><p>Here, you can see the answer generated from ChatGPT is wrong. It couldn’t understand the problem clearly.</p><h3>DeepSeek (Deepthink R1)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/766/1*zmd2j8LMvhXvdYwavQUClw.png" /></figure><p>This is the correct answer</p><p>Gemini (2.0 Flash)</p><p><em>Here’s how to solve the problem:</em></p><ol><li><strong><em>Find the sum of the numbers from 1 to 8:</em></strong></li></ol><p><em>The sum of numbers from 1 to n is given by the formula n(n+1)/2. In this case, n = 8. Sum = 8(8+1)/2 = 8(9)/2 = 72/2 = 36</em></p><p><strong><em>2. Let’s use variables:</em></strong></p><ul><li><em>Let “T” be the sum of the numbers covered by triangles.</em></li><li><em>Let “S” be the sum of the numbers covered by squares.</em></li><li><em>Let “C” be the number covered by the circle.</em></li></ul><p><strong><em>3. Write the equations:</em></strong></p><ul><li><em>We are given that T = S</em></li><li><em>We are also given that C = S/4</em></li></ul><p><strong><em>4. Express everything in terms of T:</em></strong></p><ul><li><em>Since T = S, we can rewrite C = S/4 as C = T/4</em></li></ul><p><strong><em>5. Use the total sum:</em></strong></p><ul><li><em>The sum of all the numbers is 36, so T + S + C = 36</em></li></ul><p><strong><em>6. Substitute and solve for T:</em></strong></p><ul><li><em>Substitute S = T and C = T/4 into the equation: T + T + T/4 = 36</em></li><li><em>Combine the terms with T: (4T + 4T + T)/4 = 36 9T/4 = 36</em></li><li><em>Multiply both sides by 4: 9T = 144</em></li><li><em>Divide both sides by 9: T = 16</em></li></ul><p><strong><em>7. Find C:</em></strong></p><ul><li><em>C = T/4 = 16/4 = 4</em></li></ul><p><strong><em>8. Find the sum of T and C:</em></strong></p><ul><li><em>T + C = 16 + 4 = 20</em></li></ul><p><strong><em>Therefore, the sum of the numbers covered with the triangles and the circle is 20.</em></strong></p><p><strong><em>The correct answer is © 20.</em></strong></p><p>The steps taken to solve this are a bit long but produce correct answer.</p><h3>CoPilot</h3><p>Output</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/560/1*wh3-m9d-Uqi_-JSk1qzxTw.png" /></figure><p>The answer is correct and shows some details.</p><h3>Final words</h3><p><strong>ChatGPT </strong>(GPT-4o model) struggled a bit with this problem and missed the correct results, which showed some areas that need improvement.</p><p>On the other hand, the newest <strong>DeepSeek </strong>(Deepthink R1) came through and nailed the answer, proving it has some solid problem-solving skills.</p><p>Gemini (2.0 Flash) took a thoughtful and detailed approach, leading to the right answer.</p><p>Also, <strong>CoPilot </strong>did a great job by identifying the correct solution, showcasing its talent for tackling similar math challenges.</p><p>In conclusion, while not all AI models demonstrated the same level of accuracy and clarity, some models like DeepSeek , Gemini, CoPilot showed strong problem-solving capabilities, proving that AI can effectively tackle math problems designed for young students. This comparison underscores the varying levels of proficiency among AI models and points to continuous advancements in AI’s ability to solve complex problems.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e683e30b5950" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[STOP and Fact-Check: The Dark Side of Generative AI.]]></title>
            <link>https://ambahera.medium.com/stop-and-fact-check-the-dark-side-of-generative-ai-9c66db333cf8?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/9c66db333cf8</guid>
            <category><![CDATA[ai-ethics]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[content-strategy]]></category>
            <category><![CDATA[digital-literacy]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Tue, 04 Feb 2025 08:01:34 GMT</pubDate>
            <atom:updated>2025-02-04T08:01:34.837Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ov0tLCZ9-CKfSVljUg0xrA.png" /></figure><p>A friend of mine, who has over 20 years of experience in IT, recently shared a following travel blog article with me.</p><p><a href="https://www.travelandtourworld.com/news/article/thailand-japan-south-korea-indonesia-vietnam-india-sri-lanka-singapore-cambodia-here-are-the-10-countries-to-travel-for-generation-beta/">https://www.travelandtourworld.com/news/article/thailand-japan-south-korea-indonesia-vietnam-india-sri-lanka-singapore-cambodia-here-are-the-10-countries-to-travel-for-generation-beta/</a></p><p>After reading just the first few lines, I recognized it as completely AI fabricated, as I have been studying the rise of artificial intelligence for the past couple of years. However, I believe that the general audience often lacks the knowledge to identify AI-generated content.</p><p>So, the purpose of this article is to educate the general audience on identifying AI-generated content to avoid falling into AI-generated frictional articles.</p><h3>Red Flags</h3><ul><li>The article discussed babies born in 2025 as if they were already travelling today. <em>AI doesn’t know 20 days old (At the time of that article published) infants can’t travel.</em></li><li>As per my experience, it described luxury infant care facilities in regions where even basic medical care can be challenging to find.</li><li>Detailed recommendations included nonexistent baby-friendly amenities.</li><li>Suggested travel itineraries ignored crucial factors like medical infrastructure and emergency services</li></ul><h3><strong>Why This Happens</strong></h3><p>In order to understand why this happens we need to understand how AI works.</p><ul><li>AI creates content by predicting the words that are most likely to follow one another based on what it has learned from its training data. It doesn’t actually “know” if something is true; it simply tries to put together a text that sounds logical and makes sense.</li><li>Sometimes, if AI doesn’t have all the details on a subject, it may make educated guesses based on patterns. This can lead to generalizations or claims that might not be entirely accurate.</li><li>AI likes to be engaging and creative with its language, and this sometimes leads to imaginative or exaggerated statements.</li></ul><h3>How to identify AI-generated content.</h3><p><strong>Repetitive or Generic Language</strong><br>AI often reuses phrases and provides high-level explanations without real in-depth details.</p><p><strong><em>Example</em>:</strong> A travel article described “luxury infant care facilities” in remote locations. However, it doesn’t provide any real details about where or how these services existed.</p><p><strong>Too Perfect but Awkward Flow</strong><br>AI-generated content is grammatically correct but can feel robotic or unnatural.</p><p><strong><em>Example</em>:</strong> A guide on family travel used idealized language, making every destination sound <em>perfect</em> — yet ignoring real concerns like safety, infrastructure, and accessibility.</p><p><strong>Misinformation &amp; Made-Up Facts</strong><br>AI can confidently present completely false details as facts.</p><p><strong><em>Example</em>:</strong> An AI-generated article recommended baby-friendly travel in some countries, despite these destinations lacking the necessary infrastructure. This could mislead parents into unsafe travel decisions.</p><p><strong>No Personal Touch</strong><br>AI lacks human experiences, emotions, and real-world anecdotes, making content feel impersonal.</p><p><strong><em>Example</em>:</strong> A travel guide suggested “immersing in local culture” but didn’t share any actual traveller experiences or firsthand insights — just vague, generic advice.</p><p><strong>Keyword Stuffing &amp; SEO Overload</strong><br>AI-generated content sometimes repeats certain words excessively, making it sound unnatural.</p><p><strong><em>Example</em>:</strong> “Family-friendly” was used over and over in an article, even when it didn’t make sense in context — likely because the AI was optimizing for search engines rather than readability.</p><p><strong>No Credible Sources</strong><br>AI often generates claims without citing real references. If something sounds suspicious, verify it with trusted sources.</p><h3>Why this is Dangerous</h3><p>Think about the ripple effects</p><ul><li>Travellers making non-refundable bookings based on fabricated information</li><li>Local businesses/Hotels facing guests expecting nonexistent services</li><li>Frustrated travellers will add bad reviews which will damage the reputation of destinations, hotels and travel service providers.</li><li>Most importantly, potential health and safety risks for vulnerable young travellers</li><li>Other articles will use this information for fact-checking.</li></ul><p>Ex: Perplexity used this fabricated article as its source.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/795/1*CqWc1rY4JoBoyjrs7mc9xA.png" /></figure><h3>Moving Forward Our Responsibly</h3><p>The rise of AI content creation won’t slow down. But we need to step up to -</p><ul><li>Set standards for responsible AI content use</li><li>Protect vulnerable audiences from misinformation</li><li>Balance innovation with safety and accuracy</li><li>Share lessons learned with our professional community</li></ul><h3>Your Turn</h3><p>Have you encountered similar issues with AI-generated content in your field? Share your experiences in the comments. Your insights could help others avoid similar pitfalls.</p><p>#ArtificialIntelligence #ContentStrategy #DigitalSafety</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9c66db333cf8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Can We Automate the Testing of AI-generated content? Part 3 (Key Evaluation Metrics)]]></title>
            <link>https://ambahera.medium.com/can-we-automate-the-testing-of-ai-generated-content-part-3-key-evaluation-metrics-68b9fb416988?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/68b9fb416988</guid>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[software-testing]]></category>
            <category><![CDATA[test-automation]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Tue, 21 Jan 2025 08:02:11 GMT</pubDate>
            <atom:updated>2025-01-21T08:02:11.966Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UoUncsdZnD-9qj6s16-_Ng.png" /></figure><p>Before we get into more details, let’s discuss how we test AI output today.</p><h3>How We Test AI Today (What’s Working and What Isn’t)</h3><p>Currently, we rely on three main strategies:</p><ol><li><strong>Human Reviews (Human-in-the-Loop Validation) </strong>Having QA Engineers check AI outputs sounds great in theory, but imagine trying to review millions of responses. Also, we’re all biased in our own ways, and what looks good to one reviewer might not satisfy another.</li><li><strong>Setting Boundaries (Guardrails for Prompts) </strong>In some approaches we try to control AI by limiting what it can say. While this helps avoid disasters, it’s like putting training wheels on a racing bike — you might prevent crashes, but you’re also limiting actual capability.</li><li><strong>Safety Nets (Fail-Safe Functions) </strong>Building automatic stops and checks helps catch obvious problems. However, these systems often miss the subtle stuff — like when an answer is technically correct but misses the point entirely. If you are a frequent AI user you might be familiar with this issue. ;)</li></ol><h3>Quick Recap</h3><p>In the previous article, we discussed implementing a simple Model-Based Testing framework that uses AI/ML models to validate AI outputs.</p><p><a href="https://ambahera.medium.com/can-we-automate-the-testing-of-ai-generated-content-part-1-0c7843e96543">Can We Automate the Testing of AI-generated content? — Part 1</a></p><p>But using that framework we can only validate the similarity of the answers. However, checking just similarity is not enough to validate AI generated content.</p><p>There are Key Evaluation Metrics that help us validate the quality of the output.</p><h3>Key Evaluation Metrics</h3><p>When checking AI responses, consider these key areas:</p><ol><li><strong>Does It Answer the Question? (Answer Relevancy)</strong></li></ol><p>Just like in a conversation, responses need to actually address what was asked, not just sound smart.</p><p><strong>2. Does It Use Context Properly? (Contextual Relevancy)<br></strong>Good responses show understanding of the bigger picture and use relevant background information.</p><p><strong>3. Is It Accurate? (Contextual Relevancy)</strong></p><p>Checking whether the specific details and facts line up with reality.</p><p><strong>4. Is It Complete? (Contextual Recall)</strong></p><p>Making sure nothing important gets left out of the response.</p><p><strong>5. Does It Make Things Up? (Hallucination Detection)</strong></p><p>Catching when AI invents information that wasn’t in the source material.</p><h3>Finding Better Solutions</h3><h3>What Modern Testing Needs</h3><p>We need testing that can:</p><ul><li>Work with unpredictable outputs</li><li>Keep up with rapid AI growth</li><li>Give us clear quality measurements</li><li>Run smoothly without constant babysitting</li></ul><h3>The Path So far and forward on this article series</h3><p>So, in order to effectively automate AI testing, we need approaches that-</p><ol><li>Accept non-deterministic behaviour while ensuring quality (Achieved with Spacy)</li><li>Provide quantifiable metrics for evaluation (To Do)</li><li>Scale efficiently with increasing AI deployment (To Do)</li><li>Operate autonomously while maintaining reliability (Achieved with Spacy)</li></ol><p><strong>The subsequent articles in this series will explore:</strong></p><ol><li>Implementation of model-based testing using best available tool to automate the metric evaluation</li><li>Integration strategies for continuous testing</li><li>Best practices for AI quality assurance</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=68b9fb416988" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Can We Automate the Testing of AI-generated content? — Practical Example]]></title>
            <link>https://ambahera.medium.com/can-we-automate-the-testing-of-ai-generated-content-practical-example-d124e99df52d?source=rss-e2edbe6eb3e4------2</link>
            <guid isPermaLink="false">https://medium.com/p/d124e99df52d</guid>
            <category><![CDATA[test-automation]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[software-testing]]></category>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[software-quality-engineer]]></category>
            <dc:creator><![CDATA[Chamila Ambahera]]></dc:creator>
            <pubDate>Tue, 17 Dec 2024 07:01:58 GMT</pubDate>
            <atom:updated>2024-12-17T07:01:58.129Z</atom:updated>
            <content:encoded><![CDATA[<h3>Can We Automate the Testing of AI-generated content? — Practical Example</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mV9TinDuIra_untI27KV3A.png" /><figcaption>Automate AI Generate Content</figcaption></figure><p>As discussed in our previous article, we have some challenges in automating AI-generated content.</p><p>Link to the previous article.</p><p><a href="https://ambahera.medium.com/can-we-automate-the-testing-of-ai-generated-content-part-1-0c7843e96543">Can We Automate the Testing of AI-generated content? — Part 1</a></p><ol><li><strong>Non-Determinism</strong>: AI doesn’t always give the same result for the same input, making it hard to predict.</li><li><strong>Data Issues</strong>: To perform well, you need to ensure that the AI is trained on good, unbiased data.</li><li><strong>Bias</strong>: It’s important to test the AI for biases to ensure it works fairly across all groups.</li><li><strong>Hard to Understand</strong>: AI’s decision-making process can be difficult to interpret, especially when it’s not transparent.</li><li><strong>Constant Learning</strong>: Unlike traditional systems, AI is always learning and evolving, which means testing is an ongoing process.</li></ol><h3>Practical Example</h3><p>In order to make this example simple, I planned to develop a test automation framework for AI content validation.</p><p>For this project, we are llma model which is hosted on Cloudflare and Spacy Python library for content validation.</p><h3>LLaMA (Large Language Model Meta AI)</h3><p>LLaMA is a series of powerful <strong>language models</strong> developed by Meta (formerly Facebook). What makes LLaMA special is that it’s designed to be <strong>efficient</strong> and lightweight, meaning it performs really well without needing tons of resources. It’s perfect for tasks like <strong>text generation</strong>, <strong>summarization</strong>, and other language-related activities.</p><ul><li><strong>Why It’s Great</strong>: It’s efficient, research-friendly, and works well for a variety of tasks.</li><li><strong>Where You’ll See It</strong>: Chatbots, content creation, and language processing tools.</li></ul><h3>spaCy</h3><p>spaCy is a super popular <strong>NLP (Natural Language Processing) library</strong> for Python. It’s fast, intuitive, and perfect for real-world applications. With spaCy, you can easily handle tasks like <strong>text analysis</strong>, <strong>named entity recognition (NER)</strong>, and <strong>part-of-speech tagging</strong>. Plus, it integrates smoothly with transformer models if you want even more advanced capabilities.</p><ul><li><strong>Why It’s Awesome</strong>: It’s fast, easy to use, and ready for production.</li><li><strong>Where You’ll See It</strong>: Text processing, data analysis, and building smart NLP tools.</li></ul><p>If you are using PyCharm you can install most of the requirements using that.</p><pre>1. Python 3.8 or higher<br>2. Pip (Python package installer)<br>3. Cloudflare free account <br>   Follow this guide to setup Cloudflare account and get API token to execute this project.<br>   https://developers.cloudflare.com/workers-ai/get-started/rest-api/<br>4. A supported AI/ML model (LLMA)<br>5. Follow this guide to install Spacy package.<br>   https://spacy.io/usage/models#quickstart</pre><p>For more details check the readme file in the GitHub repo.</p><p><a href="https://github.com/chamiz/ai_test_framework">GitHub - chamiz/ai_test_framework: AI Test Framework is a robust and flexible tool designed to simplify and streamline the testing of AI and machine learning (ML) models.</a></p><p>You can change the prompt and expected results in the test_data.json file.</p><p>Ex:</p><pre>[<br>    {<br>        &quot;prompt&quot;: &quot;Write a short story about a llama that goes on a journey to find an orange cloud.&quot;,<br>        &quot;expected_output&quot; : &quot;\&quot;\&quot;\nIn the heart of the Andes, a curious llama named Lola gazed up at the sky and noticed an unusual cloud with a vibrant orange color. Captivated by this strange sight, Lola decided to embark on a journey to find the source of the orange cloud.\n\nLola lived on a farm with her companion, an old alpaca named Alberto. Together, they had explored the nearby mountains, but this cloud was unlike anything they had ever seen. Alberto, sensing Lola&#39;s excitement, agreed to join her in the adventure.\n\nTheir journey took them through rocky paths and flowing streams, and along the way, they encountered various creatures. A group of playful monkeys cheered them on, while a wise condor offered them advice: \&quot;Trust in your curiosity, for the orange cloud reveals itself to those who seek it.\&quot;\n\nAs the sun set, Lola and Alberto arrived at the edge of a valley where they could finally see the orange cloud glowing like a jewel in the sky. \n\nThe two companions approached the cloud, which was no longer just an optical illusion but seemed to hum with a gentle energy. As they reached it, the cloud began to take shape into a mythical dragon that circled around them. The dragon spoke in a deep voice, \&quot;Welcome, brave souls. Your journey shows your wonder and courage. I grant you one wish each.\&quot;\n\nLola and Alberto, sharing a glance, spoke together, \&quot;We wish to share our love of adventure, so others may also discover the magic we have found.\&quot;\n\&quot;\&quot;\&quot;\n&quot;<br><br>    },<br>    {<br>        &quot;prompt&quot;: &quot;Describe the benefits of AI in healthcare.&quot;,<br>        &quot;expected_output&quot;: &quot;The integration of Artificial Intelligence (AI) in healthcare has brought profound changes to the industry, improving patient outcomes, enhancing efficiency, and reducing costs. Some key benefits of AI in healthcare include:\n\n1. Personalized Medicine: AI analyzes patient data to create tailored treatment plans.\n   \n2. Early Disease Detection: AI-driven diagnostic tools detect diseases early, leading to better outcomes.\n\n3. Clinical Decision Support: AI systems assist clinicians by providing real-time insights, reducing medical errors.\n\n4. Automation of Administrative Tasks: AI can manage tasks like scheduling, billing, and patient records, allowing clinical staff to focus on patient care.\n\n5. Enhanced Patient Engagement: AI-powered chatbots and assistants help patients manage their health and interact with healthcare providers.\n\n6. Rare Disease Detection: AI identifies rare diseases and helps develop treatments that traditional methods may miss.\n\n7. Telemedicine and Remote Monitoring: AI supports virtual care, expanding healthcare access to underserved regions.\n\n8. Medical Imaging Analysis: AI analyzes images like CT scans and MRIs to detect abnormalities and aid diagnosis.\n\n9. Accelerated Research and Development: AI speeds up medical research, helps identify potential new treatments, and optimizes clinical trials.\n\nIn summary, AI&#39;s role in healthcare is transformative, improving patient care, enhancing efficiency, and reducing costs. As AI continues to advance, new applications in healthcare will emerge, further enhancing the industry.&quot;<br>    }<br>]</pre><p>Initially, when you change the prompt you will get mismatches.</p><p>You can adjust the validation accuracy in the validators.py file.</p><pre>threshold = 0.8  # 80% similarity threshold</pre><h3>Next steps</h3><p>Integrate other AI content validation</p><ul><li><strong>SentenceTransformers</strong> for semantic similarity.</li><li><strong>Transformers</strong> (e.g., BERT) for embedding-based comparisons.</li><li><strong>Gensim</strong> for topic and vector-based similarity.</li></ul><p>If you want to check the next articles of this series consider subscribing.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d124e99df52d" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>