<?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 Mikhail Semynin on Medium]]></title>
        <description><![CDATA[Stories by Mikhail Semynin on Medium]]></description>
        <link>https://medium.com/@michael_se?source=rss-86bd8b7a418a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*OKFGvNwR-dcuaQMCVFUlkQ.jpeg</url>
            <title>Stories by Mikhail Semynin on Medium</title>
            <link>https://medium.com/@michael_se?source=rss-86bd8b7a418a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 10:33:55 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@michael_se/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 Cost of Error in the Age of Automation]]></title>
            <link>https://medium.com/@michael_se/what-can-a-mistake-cost-in-the-age-of-automation-03d73544bd8b?source=rss-86bd8b7a418a------2</link>
            <guid isPermaLink="false">https://medium.com/p/03d73544bd8b</guid>
            <dc:creator><![CDATA[Mikhail Semynin]]></dc:creator>
            <pubDate>Tue, 09 Sep 2025 18:01:13 GMT</pubDate>
            <atom:updated>2025-09-12T18:00:10.651Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*73EGnFPbmKgSe87UAwKStg.png" /></figure><p>We all know that now is a difficult period in the IT industry. Companies are fiercely competing, making deals and compromises with themselves, automating absolutely everything they can reach. And those who have been working in the QA industry for a long time have probably seen at least once how compromises within a company greatly limit testing.</p><p>But is the reduction in quality worth the risks it can lead to? Of course, in each specific situation, this is a question of risk management and the people who manage this risk. But now I want to remind you of one old but striking case when a critical error could not be detected in time and it cost <strong>more than $460 million.</strong></p><p>In this article, I will talk about a large stock broker Knight Capital Americas LLC and the events of 2012 described in <a href="https://www.sec.gov/files/litigation/admin/2013/34-70694.pdf">the report of the U.S. Securities And Exchange Commission</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8KEFW1E0K8j9pG4I7wNLeA.png" /></figure><p><strong>Here’s what happened</strong>:<br>- On August 1, 2012, the original Power Peg code was replaced with the new SMARS. However, 1 of the 8 servers was not updated.<br>- The update was performed without proper testing.<br>- At 08:01 AM, the system started sending so-called BNET reject emails.<br>- The company’s employees did not have procedures for responding to such messages and 97 failure messages were ignored.<br>- From 9:30 AM, the system began sending erroneous orders for a total of $6.65 billion.<br>- Having discovered the problem, the employees deleted the SMARS code that had just been delivered.<br>- No one tested this situation and it turned out that deleting SMARS would activate Power Peg. Now all the company’s servers began sending erroneous orders and the situation worsened.<br>- In 45 minutes, orders for $6.65 billion were sent and about 4 million transactions were executed, and the company’s loss amounted to more than $460 million.</p><p>A story worthy of a movie, but let’s look at the U.S. Securities And Exchange Commission report and understand the reason for this error.</p><p><strong>So the first problem that caused the events is Technical Debt.</strong></p><p>Of course, this problem could have been avoided, but here’s what happened:<br>- Almost 9 years before the events, the company stopped using the old Power Peg code on its servers. However, it was still possible to execute this code.<br>- About 7 years before the events, this code was partially modified. But someone decided that this change did not need to be tested.<br>- When implementing the new code in 2012, Power Peg was accidentally activated, since no one tested what would happen if the new SMARS code was removed.</p><p><strong>The second problem is Processes.</strong></p><p>I always say that a large company with a lot of money is forced to live on the basis of processes, otherwise the chances that it will not survive increase. This is how it was at Knight Capital:<br>- There were no formalized procedures for deploying code. One technician copied the code to 7 servers out of 8, without checking with colleagues.<br>- There were no processes for testing unused code.<br>- The documentation was incomplete and inaccurate, which prevented the risks from being identified in advance.</p><p><strong>The third problem is poor communication between the business and the IT department.</strong></p><p>This is how it was at Knight Capital:<br>- The employees received almost a hundred notifications, but did not know what to do and decided that it was just a formality.</p><p>This report shows that quality is always not just about one specific line of code or application function. Quality is a complex issue and must be addressed accordingly.</p><p>Can all errors lead to such huge losses — of course not.<br>Does anyone want to be next on the list — of course not.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=03d73544bd8b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Quality Assurance Engineers are not needed]]></title>
            <link>https://medium.com/@michael_se/quality-assurance-engineers-are-not-needed-ab5fca9de537?source=rss-86bd8b7a418a------2</link>
            <guid isPermaLink="false">https://medium.com/p/ab5fca9de537</guid>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[testing]]></category>
            <category><![CDATA[qa]]></category>
            <category><![CDATA[scrum]]></category>
            <dc:creator><![CDATA[Mikhail Semynin]]></dc:creator>
            <pubDate>Sat, 03 Feb 2024 10:45:04 GMT</pubDate>
            <atom:updated>2024-02-03T10:45:04.870Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x5CIsAz8aOGHEUQNr__D5Q.jpeg" /></figure><p>In this article, I intend to debunk one of the most popular misconceptions in the world of Agile development, namely the assertion that having testers on the team is not mandatory, as developers are perfectly capable of conducting testing on their own. I believe that such an approach may be justified in certain exceptional situations, but in most cases, it will prove to be ineffective.</p><p>The IT industry is constantly evolving and transforming every year. However, you can still hear opinions from developers or company managers that testers are not really necessary. After all, developers write code and, of course, check how it works. In addition, they write unit tests that often perform regression testing, which is often done by testers. And to some extent, this opinion is correct. But if you delve deeper into what QA is and what specialists in this field do, a whole new world may open up. Let’s explore this in more detail.</p><p>First and foremost, a development team consists of diverse individuals and roles working together to create a product that must meet customer needs. Each team member is responsible for quality, collaborating with others and learning from their mistakes for mutual improvement. And here, you may wonder why there is a need for separate testers when all other team members are already responsible for the quality of their work. And you are partly correct.</p><figure><img alt="let’s deploy to production without testing" src="https://cdn-images-1.medium.com/max/1024/1*882Ty63eF6qs12H6LNCu-g.jpeg" /><figcaption>let’s deploy to production without testing</figcaption></figure><h3>How They Live in Agile</h3><p>Many companies have long been developing products using the Agile methodology, so let’s take a look from that perspective.</p><p>Abstractly, it looks like this. Thanks to the information received from analysts, at the beginning of each iteration, developers start writing code, and testers start creating test cases. After developers hand over a completed module to testers, they begin testing according to prepared test scenarios, conducting various types of testing, including regression testing. Sometimes it’s a simple web form, but often it’s much more complex.</p><p>A QA engineer contributes a unique perspective to the team, sharing knowledge about testing and the product, closely collaborating with developers to coordinate testing strategies, working with business stakeholders to create acceptance tests, and notifying the start of user testing. QA engineers have been doing significant work in the application development industry for many years. They must understand how microservices, Kafka, Web Socket, REST, SQL, Docker, Git, and much more work.</p><p>Yes, a Scrum team is a self-organizing collective that may include one or more testers working closely with developers, designers, Scrum masters, product owners, analysts, and DevOps engineers. Such a team is always ready to respond to any risks and adapt to changes; it is bold and effective.</p><p>Of course, Scrum does not segregate people within the team; everyone is equal and performs vital functions for the product’s success. This is crucial because the product’s success depends on many factors, such as development speed and product quality.</p><p>Testers are individuals whose job is to scrutinize everything critically. In my view, their work assumes that there are bugs without a doubt. And in reality, this is one of the principles of testing with which I simply agree: successfully passed test cases indicate that there are no bugs where we looked for them, but statistically, they are bound to be where we haven’t looked yet.</p><p>QA engineers ask questions and evaluate the product, taking into account customer expectations, testing not only finished features but also documentation at various levels. QA engineers are at the center of everything, like spiders weaving a web over the entire team; they sense fluctuations and adapt with the team. Of course, we’re not talking about beginners; we’re talking about experienced professionals. An excellent tester can significantly influence the results.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lY8qclUCtEPGqj6bw7esAA.jpeg" /></figure><p>It’s true that even in experienced Scrum teams, there are people who find QA engineers annoying with their questions, suggestions, and criticism. But it’s like using both the stick and the carrot; someone has to play this role for the project’s effective growth.</p><p>But let’s return to the initial thesis that “developers can test on their own.” Let’s be honest here, developers usually prefer to develop code rather than write thousands of unit tests and, even more so, spend days testing their own features. Let’s also be honest that testers find bugs, often at different testing levels, indicating the need for QA engineers.</p><p>And this is just the tip of the iceberg that we see as a result of the work of QA engineers. In reality, QA can be divided into three parts:</p><ul><li>Testing and QC</li><li>QA</li></ul><p>Here, it’s essential to understand that QA engineers are not just involved in testing and controlling the current product quality. They must also help the team and the company build effective quality assurance processes, and these processes involve everyone. And it is the QA engineer who is most effective at critically assessing not only a feature but also the team’s and the company’s processes.</p><p>But let’s be honest in the opposite direction as well. There are indeed rare cases where QA engineers are not needed. A young startup creating an MVP to test a hypothesis is a suitable example. No one knows if the world needs such a product, and there’s no point in making it perfect.</p><h3>Conclusion</h3><p>QA engineers are needed by those who are in it for the long run; it’s a story about product quality and user satisfaction. Look at what successful startups do. They hire experienced QA engineers primarily to establish processes. Testing by experienced QA engineers is undoubtedly important, but predictability and quality assurance processes are equally crucial.</p><p>QA engineers may play a less visible role at times, but they are undoubtedly just as important as any other team member, and together, they can achieve remarkable results. <br>Thank you for reading.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ab5fca9de537" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>