<?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 Alexander Turkhanov on Medium]]></title>
        <description><![CDATA[Stories by Alexander Turkhanov on Medium]]></description>
        <link>https://medium.com/@alex-turkhanov?source=rss-4953e9190086------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*jPiNxi1GEikBg8Jz0J8DcQ.jpeg</url>
            <title>Stories by Alexander Turkhanov on Medium</title>
            <link>https://medium.com/@alex-turkhanov?source=rss-4953e9190086------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 02:48:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@alex-turkhanov/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[Webinar “Personalized Semantic Search using the Digital Systems Engineering Process Model”]]></title>
            <link>https://alex-turkhanov.medium.com/webinar-personalized-semantic-search-using-the-digital-systems-engineering-process-model-943331344cdd?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/943331344cdd</guid>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Thu, 31 Aug 2023 15:11:27 GMT</pubDate>
            <atom:updated>2023-08-31T15:11:27.556Z</atom:updated>
            <content:encoded><![CDATA[<h3>The problem with staffing an engineering project</h3><p>A few months back, the German chapter of INCOSE delivered us the digital system engineering process model. And they proposed a specific use case for that. But I believe it was the most apparent use case, not the most useful. They proposed the model as a device that helps you navigate through the System Engineering Handbook instead of inconveniently looking for necessary information in a paperback version. But when I was looking at it, I could not help but think that this is a semantic core for the system engineering body of knowledge. And what can you do with the semantic core? You can build a search and personalization engine. And we at Applied Knowledge Systems just decided that we should totally build that engine. The only question was what this search and recommendation engine should look like and what we are looking for. Which also was obvious. We need to look for system engineering skills and for people who have these skills.</p><p>How does one do it? To me, it looks like this. For example, we have a system engineering management plan on the input. It’s a part of project planning. And once we have a plan, we can look for people with the skills we need to execute this plan. Should be as simple as that — plans on the input, people candidates on the output. The reality, though, according to the process model presented in the Human Resource Management process of the handbook, has a cumbersome, very tricky, and very time-consuming structure. It starts with project planning, and as a result, we have a Project Portfolio and Project Human Resources Needs. Then, we identify, develop, and acquire and provide skills. On outputs, we have a human resource management plan, qualified personnel, a human resource management report, and a human resource management record. Nothing that looks like the process I have in mind. It’s really time-consuming, and we can eliminate a lot of waste in this area, precisely thanks to the systems engineering process model and role-based search capability we built.</p><p>If we are successful, INCOSE as an organization or for Indra or other engineering companies that have system engineering as a part of their operations can streamline their human resource management process, and that will not require getting tons of approvals, building connectors, and changing informational security profiles because we built privacy-first AI assistant that does not send your data outside the already approved network or your workstation. I will explain how it works and show you the prototype. My ask from you is simple. So far, we have done all the difficult jobs of building the platform and data processing pipeline. We can now build a product that actually solves your problems. It’s a fun task you can all participate in.</p><h3>How searching for skills should look</h3><p>I want to thank the Spanish chapter of INCOSE and specifically Anabel Fraga, who helped me a lot in organizing this webinar. I also want to thank Alexander Mikhalev, the CEO and developer of Applied Knowledge Systems, who spent a lot of time building the Terraphim search engine. I am both product manager, head of product, and project manager in this small company, as it usually goes in start-ups. But in corporations, we also execute several roles directly related to this webinar’s subject.</p><p>When you search for something, you do it differently from the perspective of a CEO or from the perspective of a project manager. If I am a project manager preparing an outline project plan, and I look for some topic, I need a clear and concise explanation that I can use to formulate project tasks that will go well with the project management office. If I am a CEO and use the exact keywords in the query, for example, “Forms of contracting for product companies,” I most definitely need practical guidelines and actual contract templates, preferably with comments from the lawyer for a specific jurisdiction. But search engines do not support that role distinction; maybe only Perplexity provides a decent experience. But even with it, you still need to think about how you will search as a CEO or how you will search for something as a project manager, and it just supports further queries when you know what you are looking for and why.</p><p>Imagine such a scenario. We have a project plan a system management plan, and I’ll take the actual task description from the Notion workspace of our Innovate UK project, where we have forecasted time sheets, registries, and reports, as well as deliverables and milestones. I put this task description to my personal assistant. It investigates the system engineering digital process model and hints to me that OK, I need project management skills to accomplish that task.</p><p>It gives me a description of this skill, so I understand what exactly stands behind it. This is really a machine-generated project description of that task. This is not mine, and it’s actually good. OK, I read it, and it seems legit. So this is a proposal to collaborate generated from the task, and I respond OK, I approve this skill description for my project. What should be the next step? I think you should add some supporting skills descriptions because it doesn’t go alone. Project management skills also require risk management, stakeholder management, communication skills, and problem-solving skills. Again, these are all skills that have been proposed by the machine based on that proposal to collaborate. She generated them, and I said, “OK, can we be more specific?” And it goes to the curated WAND taxonomy. WAND company has hundreds of them and elaborates on the skills descriptions using these taxonomies. We have a project management taxonomy containing more than 1000 terms. I ask myself — what are the areas of project management I am most interested in to manage this task? It’s not completion, it’s not documents and records, it’s not execution, it’s planning and design. I selected it and realized I don’t need design; I need planning and methodology. Specifically, I need a critical chain method. Using WAND taxonomy, I substituted abstract and broad project management skills with narrow and specific “critical chain” skills.</p><p>It does the same with risk quantification, stakeholder analysis, and communication management plans. So now, from a generic task description that is just three lines of text, I have a pretty concrete project proposal with pretty concrete skills. I want a candidate to have all of them. And I say, ok, find me the person that can be a good fit. It goes into my contact list, creates a new page in my Notion workspace for this project, and sends me the link, and I see this resume. Again, this is a machine-generated resume, and it is pretty good. I look at the timestamp of the resume and understand that it was generated one minute ago. The person whose skills are described in this resume has spent not a minute, no time at all, zero effort to write it specifically for this particular task proposal, the machine has generated it, and it contains actual information because this person has a lot of personal GitHub commits, a lot of Notion documents and Discord messages. We have a lot of information about this person, and the machine just fetched it and generated a resume from the original working products the candidate has created.</p><p>And what’s the beauty and benefits of generating proposals to collaborate and generating matching resumes from the actual work products instead of the standard human resources process I described above, with project portfolio and human resource needs, acquiring skills, and qualified personnel at its outputs? Such an approach based on generative AI and semantic search finds better candidates for the job and removes over-secured search for usually overqualified personnel. It also supports each statement in this resume, and it is not just lines of text; it’s not just a string. Every sentence is actually a link, and I can verify the evidence in GitHub in Notion to check how valid each statement is in the resume. Is it a real thing or not? If unsure, I can ask, “Please generate a controlling question for this statement.” So if the person says that she made realistic project plans, OK, the question would be, “How did you make sure your project plan is realistic?” That’s a legitimate question. Or you can ask, “How did you attend to the project’s performance baseline?” The project performance baseline actually should be completed, so that’s how it works. That’s how I can arrange the assessment of candidates.</p><p>But what is the semantic search? How all of the above is related to the topic of the webinar? As you can see, we are looking not for strings but for the resume’s matching keywords. We are looking for confirmed skills the candidates obtained. And when I am looking for skills, I am looking for things. And that is a definition of semantic search. What’s important here is that when you have things as a result of your search, you can organize them into taxonomies.</p><h3>Context of searching for skills</h3><p>That’s quite unusual turn, one can say. What’s a taxonomy? If you go to the Amazon website, you see them on the left sidebar. If you are looking for shoes, you will see shoe taxonomies on the left side, being that taxonomy of shoe color and sizes, brands, prices, design, and purpose. It’s all organized in these hierarchical structures. Such structures are taxonomies. They significantly improve the search experience, as we know from online stores.</p><p>Do we use taxonomies when we search for talent? Not very much. Take LinkedIn, for example, where skills are listed in a plain list; you scroll them down, and no structure helps you navigate those skills. This is a consequence of not grounding skill descriptions in real-world persons. On LinkedIn, you search for strings, which are very difficult to classify, and put a taxonomy structure on them. SFIA framework or professional standards like systems engineering competency framework provide more guidance, but real professional profiles usually have more specific names for skills that are not always aligned with the taxonomies described in the standard. And anyway, a significant share of companies use plain lists to manage their skills. But we have a prominent workaround for this gap — we can use the process model as a proxy for skills.</p><p>Systems engineering has a process model that can be used as a proxy for the skills taxonomy. And this model we have in the machine-readable form. We built a demonstration that uses the system engineering digital process model to augment the systems engineering skills search. The model contains four processes: technical, technical management, organizational project-enabling, and agreement processes. Each process has activities, and obviously, each activity and process requires relevant skills. It is a hierarchical structure that is aligned very well with the systems engineering competency model as a reference catalog of professional skills. As I stated in the beginning, all the terms from this model comprise the semantic core of the systems engineering domain. That’s a sound semantic model that can augment the task of searching for skills. Let’s return to the use case I started this talk with.</p><p>If we use a systems engineering management plan as an input to the semantic search of the systems engineering skills, then such taxonomy could prove itself useful, increasing the skills search relevance. The task of searching for skills is often overlooked. For example, the famous teambuilding framework introduces five stages in the team life cycle — forming, storming, norming, performing, and adjourning. But before you can start forming any team, you need to discover who should be in it. You need to be able to search for relevant skills, generate proposals to collaborate, and, finally, onboard team members. The project manager, the chief systems engineer, or any other technical leader should bring them together first. Also, there is more to the task of staffing an organization, as one should bring team members together and stakeholders.</p><p>This task can prove to be difficult, as once the project proposal gets approved, the timeline to arrange everything, find the right people, build up connections with stakeholders, and establish proper organizational interfaces is usually tight. When they give you the money, they want results fast. That’s the way it works. And people are not always at our immediate disposal, so you need to find them quickly. The average project has hundreds of tasks, each requiring several skills, so we are talking of hundreds of skills required to complete the project. A qualified system developer or technical leader has over a hundred skills, with over a couple dozen active and 60 to 80 passive skills.</p><p>It’s impossible to know what everyone on your team does and find the best match quickly for all the tasks, issues, risk containment plans, audits, and gate reviews, especially for projects using novel technologies or during the high churn of personnel. If you look into a typical task tracker, you will find a significant portion (from my experience, I’d say at least 15%) of tasks and issues without the responsible person assigned. Also, you will find that many tasks are assigned to the same very qualified engineers, leaving them overloaded and being the bottleneck, and some other staff sometimes without much to do in the downstream development processes, especially at the beginning of the project. Finding proper and economically viable skills in the engineering organization is not a trivial task. We often tend to lean on a few highly qualified and well-compensated people.</p><p>But the problem extends far beyond just staffing the project. What if we need to find the right participants for the customer meeting? Or what if we hire a new system developer and are required to gather the interview panel and define the onboarding sequence? How do we find the people who are best fit for this job and have enough time? Because the costs of meetings and new hires with traditional means are staggering and are often well over a few thousand dollars. Is the agenda for such a meeting aligned with the competence profiles of the participants? Are they competent to make the decision? One of my favorite examples is responding to a request for a proposal or starting a new project. We usually have a very tight response timeline when we get an RFP. At best, it is a couple of weeks. That means you must find all the people you need to prepare the response in two days.</p><p>To quantify the problem of skill search, I propose you perform a small experiment with your LinkedIn profile. LinkedIn has a feature for measuring your social selling index, the SSI. This index has four components. If you are close to an average user, the lowest value components in your profile would be “Find the right people” and “Engage with insights.” It is the hardest thing for all of us. In my professional network, there are more than a thousand people, with most of them, I communicate rarely and would appreciate assistance to learn better what their skills and experience are when I get in touch with them.</p><p>That is precisely what we are building in Applied Knowledge Systems — an action-centric privacy-first personal AI assistant that can fetch you proper context depending on your next actions. You can feed it a project plan from your Notion, a product roadmap from GitHub, or a meeting invitation in Gmail, and it will generate your messages with invitations to collaborate and recommend the right people from your contacts list. Such a solution does not exist right now.</p><p>Also, we all have been facing search quality degradation for the last few years and the loss of privacy, which is of high concern to us. That’s why we are building Terraphim as an AI assistant that is entirely under your control. And here, we cannot rely on companies that ignored the privacy and degrading search for years to deliver a proper solution. Also, big companies usually build mass-market solutions that do not account for industry-specific needs and aspects. This is because such niche market segments do not justify the development costs for them and also because the users will never provide the company-specific data required to customize the solution. All tailoring that big companies can offer us comes in the form of prompt engineering, which, more often than not, is not sufficient and too time-consuming for customization.</p><h3>Pareto-best work-to-skills matching</h3><p>What can we expect to see if we successfully resolve the problem? According to Gallup reports, many people are miserable at work because they do not use their abilities to the full extent or are overloaded, precisely two things that have the same root cause of the biased work-to-skills matching process. The numbers of workplace dissatisfaction are staggering. If we can match work to skills better, we will see an improvement in workplace satisfaction as most occupied professionals will be less busy, and people who are not using their talents to the full extent, will apply more of their skills.</p><p>As I mentioned above, a task or a set of tasks usually requires several skills, so in fact, we are looking for multiple skills match, and most of the candidates cannot be equally good at all of them, which still should be fine in most of the cases. For example, I am pretty good at requirements engineering and systems architecture, though my major is lifecycle and configuration management, but for some tasks, where my level of requirements engineering is sufficient, I can be a good fit, though there are a lot of people who know it much better than me. However, they could not know configuration management, which is also required to get the job done. Such competence profiles are called “Pareto-best.” We are optimizing the search to match as many skills as possible, but giving the key skills more weight than the secondary. Again, if you use the traditional search, such an approach is impossible. All keywords in the prompt have the same importance, or you can have mandatory keywords, selected with prompt syntax, and others that the search engine can completely omit, reducing the search relevance. You cannot find and properly rank persons with requirement engineering, systems architecture, and configuration management skills, but showing more competence in configuration management first and filtering out those who do not know requirement engineering at all. The search prompt syntax doesn’t allow that. You cannot implement Pareto-best skills search with the traditional technology, but very well able to do that in Terraphim.</p><p>The textbook quality of the data we have in engineering projects makes it much more attractive to apply to engineering companies. There is a direct relationship between the dataset quality and the quality of the chatbots. You can go and check how well chatGPT answers systems engineering questions and compare that with its performance in marketing or product management fields. Answering systems engineering questions will be much more reliable because the training data are so good. The only problem is that these data are distributed across multiple sources, and we need either federated search or robust, reliable, and private interoperability between data sources.</p><p>Another aspect of the proposed solution is improving the search results’ relevance in small bits. If you have 15 minutes, you can train Terraphim to understand you better. You don’t need a month-long implementation to build the next increment. You build your own language with which you talk to an AI assistant, not studying proper prompt engineering as you need with commercial LLMs, which use their own lingo and will not understand your terms. This is just how they are built. Such lengthy implementations and complicated query languages are not aligned with our workplace reality.</p><p>Also, you want to exchange information about skill profiles. You need to map skills descriptions to some widely accepted standards to do that. In this demonstration, we use the SFIA framework, The global skills and competency framework for a digital world, and a few WAND taxonomies (project management, for example) to enrich the original systems engineering process model with more specifics and provide interoperability. The SFIA framework works miracles with chatbots. Try the Bing chat or chatGPT for yourself, and ask them a question about your job experience. For example, copy and paste a paragraph from your resume and ask what SFIA skills you have based on the text, and it will provide a very good starting point.</p><p>As all these standards for skills use controlled semantics and vocabulary, it is very easy to produce formal activity-centric models from them with different tools, and we apply Apollo 4D activity modeler because it implements the best practices introduced and developed by Mathew West and resulting models can be easily fetched by chatbots. Of course, you can use machine-readable versions of the Object Management Group standards or WAND taxonomies without spending any time on serializing text descriptions of skills into proper JSON. Once you build a formal model for some skill, for example, for the SFIA Project Management, Level 4, shown in the illustration, you can ingest it into the AI assistant, and it can generate dozens of interview questions on the subject of project management. For example, it can ask a user, “What Project Management Tools did you use to Manage Risks when you were working at Company X in 2019?” It takes the formal skill model and the user’s resume and generates a proper question, tagging it with domain terms — Project Management (SFIA skill), Project Management Tools (SFIA concept), and Manage Risks (SFIA concept). After it gets the user’s response, it transcribes it and tags it with the same domain terms. So, when you search for these keywords, you will get a match for confirmed skills that are correctly placed under the SFIA skills taxonomy.</p><h3>Role-based sematic skills search</h3><p>So, on one side of such a semantic model, you will have formal structures, taxonomies, and ontologies, and on the other, it will be a natural language response coming from subject matter experts. You can keep both particular professional lexicons in your response to the AI-generated question, which will still be searchable and discoverable because it is mapped into the controlled vocabulary of the systems engineering digital process model, WAND taxonomies, or the SFIA terms. AI assistant takes natural language input and tags it to skills taxonomies, which gives you the power of formal models for search but doesn’t require much effort to build such models.</p><p>But we went even further. Tagging the content significantly improves search relevance, but you will use several thousand formal terms from about 15–20 taxonomies in the average project. Every piece of content will be tagged with different terms from different domains. For example, the same project proposal will be tagged at least from the project management, engineering, finance controlling, and corporate compliance. Even for a small project with a team of five people, we will have mixed relevancy ranking just because the project manager, developer, and finance controller will have different preferences for how to rank search results. Different roles imply different relevancy rankings and different taxonomies to augment the model. With roles-based search, for the same search query, project managers will find one person and engineer another because they do different jobs, and thus, with an action-centric approach, they must benefit from varying recommendations.</p><p>That’s why we implemented the role relevancy mechanism in the Terraphim AI assistant. Please see this demo of the changing search relevance once the user selects different roles. In fact, Terraphim performs a two-step search — first, we define domain relevance using keywords, and then we define further actions relevancy using a role-based search mechanism. It may sound vague and complicated, but all these improvements come from your note-taking application: Notion, Fibery, Logseq, or Obsidian. We also can process GitHub repositories or Jira issue trackers. You don’t need to do much on top of what you are most likely already doing as part of your daily routine.</p><p>The Terraphim role is a combination of three things: each role has its own set of data sources, each role has its hashtags, objects, and role-specific activities, and, finally, each role has preferences in prompts and history of search results. Let’s dive into them. The combination of data sources means that I am in the father’s role; considering the class’ WhatsApp chat and educational platform as relevant sources of information and the project workspace as irrelevant, I don’t want to get any results, even well-fit, from my workspaces. And vice versa, if I am searching for something from the project manager position, I don’t want to get the results from my personal e-mail or family photos folder. If I select the role of a systems engineer, I have specific high-priority hashtags such as life cycle model, system configuration, or concept of operations, but for example, a project’s budget has low relevancy in my search. The situation changed when I switched to the project manager role — project budget became a highly relevant hashtag, as well as work breakdown structure or deliverables, but system requirements do not matter for me a lot, and that affects search results ranking. And finally, the user can approve or ignore search results, fine-tuning the assistant’s relevancy function.</p><p>And that’s it. This is the first Terraphim use case we are implementing now. We are streamlining project staffing with engineers by automating steps in between using a controlled dictionary and curated taxonomies we ingest from industry standards and the catalog of WAND company, and we are using the roles mechanism to augment the search relevancy. Thank you very much for your time; please visit our website and participate in testing the early version of the product. Help us build your personal privacy-first AI assistant that fits your daily workflow.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=943331344cdd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Terraphim — Privacy Preserving AI assistant]]></title>
            <link>https://alex-turkhanov.medium.com/terraphim-privacy-preserving-ai-assistant-1c7bd7110112?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/1c7bd7110112</guid>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Mon, 14 Aug 2023 11:12:23 GMT</pubDate>
            <atom:updated>2023-08-14T11:12:23.989Z</atom:updated>
            <content:encoded><![CDATA[<h3><a href="https://www.youtube.com/watch?v=tQECA97u1rY">Terraphim — Privacy Preserving AI assistant</a></h3><p>We are starting beta testing of our new product.</p><p><a href="https://www.youtube.com/watch?v=tQECA97u1rY">Terraphim — Privacy Preserving AI assistant — YouTube</a></p><blockquote>If you are interested, please visit our website <a href="https://terraphim.ai/">Terraphim AI</a> and fill out the contact form.</blockquote><p>Terraphim — an easy and privacy-preserving way to take control of search relevance using your Logseq.</p><p>“As a Logseq user, I routinely look for new ways to improve my flow. Why am I doing this? My graph grows every day, and while it becomes more and more practical, things in it tend to become less coherent and more challenging to find. Sometimes I review my notes and wonder how much good stuff I forgot about I have there. So, I need improvements regularly and need to make use of the information I capture.</p><p>I am a structured person, I follow processes and build my knowledge graph, but will AI assistants help me and use my knowledge graph?</p><p>Recently I especially liked AI assistants like Grammarly or Notion AI. They do a fantastic job with a single page but cannot help to improve the structure of the large enough knowledge graph. This is because all AI assistants, like chatGPT or OpenAI, ignore all markup, frontmatter, backlinks, and highlights. Everything that matters so much to the user they delete; they don’t need it. A knowledge graph is not a tree of ideas for them; it’s just a tree of strings. It’s all strings for modern AI assistants, and it undermines the semantic search’s fundamental principle “We are looking for things and not strings.”</p><p>We built a new privacy-first AI assistant — Terraphim, that does not ignore your italic, bold, highlights, and backlinks. How does Terraphim improve your flow in terms of structure and findability and make you think about things you know in an action-centric way? Being action-centric is vital to us. Life is not always an exam where you must show what you know; it is often about how you apply what you know.</p><p>Let’s have a look at a demo flow. — Terraphim can:</p><p>- Select and zoom your concepts from a background on the page so you easily remember and apply all that you already know about the topic on your go, with complete privacy.</p><p>- Search and discover in the cloud, so you can find the information you need at the right time when you need it, helping you get things done. With Terraphim, you can stop thinking about documents being lost somewhere in chats, repositories, and local and network drives. It will collect everything and find the right piece for you. <strong>And what is essential here — you are always in control of search relevance using your Logseq. You define what is right, not some corporation.</strong></p><p>- And finally, with Terraphim, you constantly improve your knowledge processing in Logseq.</p><p><strong>To learn more, visit our website. </strong>There you will also find our contacts to address your questions, or you can join the early adopters’ program or the waitlist to help us build the tool you need the most.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1c7bd7110112" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fancy product roadmap or good-old MRP II production plan?]]></title>
            <link>https://medium.com/design-bootcamp/fancy-product-roadmap-or-good-old-mrp-ii-production-plan-f24ec712fa86?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/f24ec712fa86</guid>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Fri, 16 Jun 2023 10:39:16 GMT</pubDate>
            <atom:updated>2023-06-16T22:57:54.370Z</atom:updated>
            <content:encoded><![CDATA[<p>I was listening to the <a href="https://www.youtube.com/watch?v=KhPwlIjh70I&amp;list=PLxS3M9r3X-_l2tFx0nBTba87qCpVbwM8M&amp;index=14">“Algorithmic feature prioritization, Featured Product Maker, Gabriel Steinhardt”</a> webinar and making some notes when it occurred to me that I needed to recover presuppositions hidden in the facts and logic of the report to produce genuinely valuable feedback to the author and his audience. And that kind of deep thinking (Newport, 2019) I have been trying to practice for the last five years led me to uncover one misconception that plagued product management semantics over the previous years. I am talking about “roadmap prioritization” (3,140,000 results) or “product plan prioritization” (11,900,000 results), which, as you will see, does not make sense from the point of view of control and queue management theories. And I am too old-fashioned systems engineer to abandon mathematics for marketing buzzwords. You cannot prioritize plans; you can prioritize only queues. Do not mistake causes and reasons here. Reverse causality is still an error.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j0-77LotwpwHbg-gjGUlyg.jpeg" /></figure><h3>Not every time series is a queue.</h3><p>Let us consider a couple of common examples to illustrate the main misconception. Imagine a retail chain with hundreds of stores or delivery points — Mercadona, Amazon, Walmart, Costco, or Home Depot. Each store has a line of customers or even several lines in places like Mercadona. You can manage these lines by adding cashiers, for example. Each cashier executes customer orders, and the store IT system consolidates them into the list of purchases in this store. Then the retail chain IT system puts the sales data into a single list of purchases in this retail chain. And tax agency lists all receipts or bottles of alcohol sold on the national level. Can we say that the list of records in the retail chain database is a queue, and we can manage it? No. It is just a times series, though it has some similarities with a queue — it has an order, a sequence, which we can confuse with a priority of elements in a queue. Millions of people are running sprints, but only the 20 fastest people are on Earth. But the record sheet is not a queue, and the position in it is not a priority.</p><p>Consider an airport; we have a list of arrivals on the tableau. It has up and down, and flights have a place on this list, but it is not a priority. If some flight gets delayed and you need to prioritize it, you usually do it on the departure side in the other airport, not at the side of the arrival. You see <strong>a plan</strong> on the arrivals schedule, and I cannot imagine how to prioritize it. You prioritize several departure schedules to get the plane to arrive on time. Yes, you want it to be on time so everyone gets their connection flights and takes their buses or gets the venues, but it is not a prioritization. You manage the production schedule here to meet contractual obligations, which is another discipline and conceptual framework altogether.</p><p>The same logic applies to an engineering company. It has several life cycle implementations, chains of departments, work groups, suppliers, technological partners, and supporting business functions with backlogs of stakeholder needs, system designs, regulatory requirements, and technical debt issues. Yes, <strong>each system life cycle finally results in a single flow of released features and documentation</strong>, which we can list in the same way the retail chain IT system records all the purchases, or tableau of arrivals shows planes sequence. And yes, this order matters a lot. But do not confuse the priority of getting off the plane with an arrival schedule. Do not confuse the features production plan (and this is what the product roadmap really is) with backlogs. There is a line for getting off the plane, and you can act to be the first in it, but the arrival schedule is a production plan, and you should have done something upstream to get ahead of it.</p><p>You can manage feature backlogs, but you cannot prioritize product roadmaps.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f24ec712fa86" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/fancy-product-roadmap-or-good-old-mrp-ii-production-plan-f24ec712fa86">Fancy product roadmap or good-old MRP II production plan?</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Will continuous everything ever fly alone?]]></title>
            <link>https://medium.com/design-bootcamp/will-continuous-everything-ever-fly-alone-bccb6fa53d78?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/bccb6fa53d78</guid>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[project-management]]></category>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Mon, 05 Jun 2023 05:00:47 GMT</pubDate>
            <atom:updated>2023-06-06T08:37:51.422Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*q5YCwrp7yeaTTGmu9rs5lg.jpeg" /></figure><p>I was discussing the last version of the INCOSE Systems Engineering Handbook. There is still a strong emphasis on the stage-gates model, milestones, and project management practices. And that ignited heavy critics of my opponents, who have a strong software development background and DevSecOps.</p><p>“We don’t need these obsolete concepts. We are continuous everything!” they said to me. Well, you are not. I’ll go ahead and describe my arguments here. There is one crucial reason you can never eliminate, even in the most idealized environment with the dream team, in a profitable dream organization with sufficient funds and no strings attached. And that is called the banking system.</p><p>The banking system uses transactions — you send and receive money in chunks. And you put details on each transfer — the reason, sender, and receiver. It works in a certain way because you should always be able to explain why you are sending someone some money to the regulator. These transactions are for work done, products sold, or services made that are limited in two aspects. Technology (there is a proper way to make a work, a technology), and compliance (there is a certain way to explain why one party paid another party money). You can be continuous on your technical side, but still, there is a transaction on the compliance side. You cannot pay cash for every language-fit experiment or line of code. The anti-laundering department will stop this nonsense.</p><p>And here comes the mirroring hypothesis (Colfer). Suppose you have organizational modularity (payment milestones). In that case, it will reflect on your system’s architecture and force you to use the project plan as part of risk management activities. After all, you will need money to be transferred through the banking system and enforce some modularity on your delivery cycle.</p><p>So, whether you like it or not, the transaction nature of any organization is inevitable, and so are the projects with all their structures, practices, and compliance. DevSecOps and continuous everything will never fly alone.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bccb6fa53d78" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/will-continuous-everything-ever-fly-alone-bccb6fa53d78">Will continuous everything ever fly alone?</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What to do if PDCA stops working?]]></title>
            <link>https://alex-turkhanov.medium.com/what-to-do-if-pdca-stops-working-698351fd7271?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/698351fd7271</guid>
            <category><![CDATA[data-driven-decisions]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Sun, 04 Jun 2023 08:39:30 GMT</pubDate>
            <atom:updated>2023-06-04T08:41:35.244Z</atom:updated>
            <content:encoded><![CDATA[<p>Change the system level to the higher one or, the lower one. Let’s bog down into the details.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QP2uj4BP_Ns7cinPc4Z-pQ.jpeg" /></figure><p>I am passing exams now for a Spanish driving license. And at one moment, I realized that the current PDCA cycle no longer works. I look at the data (my error sheet), and I don’t know what should I change in my driving habits and what skills should I train to score the passing grade. The data is not consistent anymore. Earlier iterations of exam simulation surfaced deficiencies I had, and I followed those data to make and implement decisions on changing my driving style. Still, the last couple of simulations gave me nothing. A mistake here and there, but nothing that can be tied to the predefined list of skills that one must obtain to pass the test successfully. I could do as many cycles as I wished, but the plan apparently doesn’t work anymore; its PDCA cycle is broken.</p><p>Obtaining a driving license here can be lengthy and costly, and my project, as any other project, has its constraints. So I went for a bicycle ride, but this time I did not go in the mountains or the beautiful fields; I took the “dirección centro.” That is absolutely wonderful that you can ride a bike here on the streets. They are remarkably safe for bicyclists and pedestrians, nothing compared to the streets of Moscow. I hung them in the early morning, looking at road signs, lanes, and youngsters, returning from the night party, crossing the streets wherever they wished, and imagining I was driving a car. And at some point, the insight came to me. I was driving a bit too fast in the car to navigate the city streets properly. When I ride a bike, it’s natural to follow all the rules, with its speed somewhere 15–20 km. So, I thought the speed limit of 30 is for drivers who already know the city very well, and they can drive that fast and still follow them, but to pass an exam, you need to go a bit under the limit, not a bit over the limit, how my instructor taught me.</p><p>Do you see what has changed here in this consideration? A system level. Passing an exam requires thinking not about a vehicle’s behavior but about the traffic flow you are in. You are a part of some train — other cars, trucks, bikes, and motorcycles around you. And road signs apply to all of them, so their maneuvers should be coordinated, and that requires following the rules, even the ones that look excessive, like very smooth acceleration, breaking, overpassing, “Yield a way” without stopping, and things like this. A car does not provide you the transportation capability, and a vehicle is an element of more extensive capability — city, national, and European transportation capabilities. And on each level, architectural aspects change; for example, achieving the EU road safety goals on the higher system levels requires many restrictions and additional rules to be applied on the level of a single car.</p><p>My PDCA cycle stopped working because I was improving the wrong process on the wrong system level. It required fixing, of course, but that’s not enough. Every driver is a trained operator who contributes to the overall safety of the transportation system, which I enjoy so much. Following the strict rules during an exam becomes certification guidelines for safety-critical system operators. And PDCA starts working again because you see your car driving process through another perspective. And it ultimately makes another line of questioning the data from the exam simulations. Not what did I do wrong when driving a car? Because there will be many things that will not make any sense. But “What did I do wrong for a traffic train I was part of?” And that question makes much more sense.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=698351fd7271" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Three layers of “products”: conceptualization framework]]></title>
            <link>https://medium.com/design-bootcamp/three-layers-of-products-conceptualization-framework-d49b4c00e09a?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/d49b4c00e09a</guid>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Tue, 30 May 2023 05:49:00 GMT</pubDate>
            <atom:updated>2023-05-31T07:44:48.432Z</atom:updated>
            <content:encoded><![CDATA[<h3>Abstract</h3><p>Using well-defined terms from systems engineering can lead to clear communication within product development teams. We put a label of “product” onto the product system of interest (SoI), service SoI, and product-as-practices-of-making-money. These three types of products produce hierarchy, a systems’ holarchy. We have capabilities, make services on top of them, and find additional monetization models besides selling things and service subscriptions. We call all these different things “products,” though various organizations build them using different kinds of plans and multiple approaches.</p><h3>Opinionated domain</h3><p>Imagine that the R&amp;D director who commands 4 product lines comes to the CEO and states that he doesn’t know how to:</p><ul><li>Establish a coherent methodology for planning and control across all product lines. He doesn’t know what milestones, phases, and typical deliverables are required; whenever the customer wants a new product, he starts planning almost from scratch.</li><li>He cannot establish organizational structure and reporting lines for departments. He doesn’t know if project management, operations, and design should report to him or another CxO.</li><li>And finally, he concludes that it is impossible to discover and follow best practices for prioritization, planning, and new product launch for development tasks. And there are no dedicated tools to do R&amp;D jobs, and they perform most of the functions in Excel, Notion, or Confluence. Texts and tables, that’s all they have.</li></ul><p>As CEO, you know what to do — you need to fire the guy who isn’t qualified and promote someone from R&amp;D to the new position. You also talk to the best people from this department and find out they are clueless. 9 out of 10 are dissatisfied with the organization and consider that the product’s success depends on a particular leader. If she is good, the product survives. If not, it will be a failure. Then you go on the market to understand if it is just your problem and find out that maybe the top 20 companies do better, but most of the market lives like that, and that’s normal. When you try to find out why things are the way they are, 20% give a weak explanation, but most are clueless.</p><p>The problem above is almost bogus. And that’s a bizarre situation, don’t you find? Every business function knows how to organize its activity. They know what to do in finance, production, operations, purchasing, and logistics. Engineers also don’t have these problems. Engineers have had locally optimized processes that are not integrated across the entire enterprise, poor alignment and coordination of the extended enterprise, or unclear roles, responsibilities, and accountability (Oehmen J. et al., 2012) ten years ago, but at least now they don’t lack the knowledge about how to build R&amp;D organization to achieve different goals.</p><p>But product organizations still have them (The 2023 State of Product Management Report, 2023):</p><p>1. Product organizations don’t know and cannot follow best practices in planning, control, and resource prioritization — nine people out of ten are dissatisfied with these business aspects and consider that success depends on the skills and aptitudes of particular professionals.</p><p>2. The industry does not have an answer to the root cause of these organizational problems (systems engineers know that very well in hard figures). Only around 20% of respondents gave some reason for the poor corporate performance. Others state the problem and don’t know how to resolve it, even in theory.</p><p>3. These problems are immaterial for other business functions. Engineers, finance controllers, and product floor people have proven business rules, best practices, and fact-based guidelines. They are not as opinionated as product management, where many gurus and working groups develop incoherent frameworks.</p><p>The research hypothesis is that there is a fundamental misunderstanding of product management concepts. And that has implications: product organization (structure, business functions, organizational interfaces, processes) is still fluid. Unstable domain boundaries and unclear definitions impede the understanding, onboarding, and career path planning and drive inefficient knowledge and information management of the product management domain. I argue that we can achieve the proper overall understanding of the product manager role only when considering how three transdisciplinary approaches work together. Namely, how system engineers, program managers, and product managers work together and what concepts and tools they operate while using the same terms for entirely different things, which causes ontological rattling during communication. This work is the first step that establishes the foundational concepts in the product category.</p><h3>The missing protocol</h3><p>Science and technology studies (STS) revealed how researchers and engineers agree on something, how they develop theories, how one approach wins over the competitors, and how the social construction of the facts works. STS gives us the know-how to produce evidence from mere and unreliable data. In other words, it explains how research protocols operate in the laboratory and design center to make a convincing narrative in a room full of decision-makers. When I looked at the ProductPlan report through the lens of STS methods, I saw this problem differently.</p><p>Let’s be specific for the sake of further argument. The report mentioned above shows that <strong>the old discussion about outcome vs. feature product development plans</strong> is still far from resolution, as 54% say that their product roadmaps focus more on outputs, and 43% say they focus on desired outcomes, with 3% choosing other options. We look at the presupposed dichotomy when we see a 100% pie cut into three portions, but here we make a mistake in ontological foundations.</p><p>That is a natural way to think about feature-based and outcome-based as opposed because product organizations oscillate between these approaches. We see that a lot. And that makes us feel that these product planning frameworks comprise classification — we are either in this box or in the other. Running is contrary to sitting, falling is contrary to flighting, and serving a customer is contrary to following instructions. It is easy to mistake <em>two different states</em> of a system for <em>two operational modes</em> of a system, but they are fundamentally different. States can overlap, like I can be both hungry and sick, modes cannot, and I cannot be both excited and depressed. And My organization uses output-oriented roadmaps as a state, not a mode.</p><p>It’s much more like I was running, then I got hungry, then I was running again. Yes, you can build a pie chart that will say what percent of the day you were running and what you were hungry for, but something will be off with it. It’s just not so apparent with questions about product road mapping techniques. <strong>And as product managers are missing the consensus-building protocol laboratory researchers utilize, they cannot uncover such hungry vs. running situations and become victims of endless debates.</strong></p><h3>Three layers of “products”</h3><p>(Kun, 2016) describes the first step for such protocol — provide clear definitions, with good examples and counterexamples. And here, you will find a big surprise, as different language communities use the term “product” with roughly three different meanings. But we can arrange these definitions into a clear hierarchy, using systems thinking and 4D ontologies (West, 2011; Partridge <em>et al.</em>, 2020; Partridge, 2021; <em>Apollo-Protocol/4d-activity-editor</em>). A product has a layered architecture if we consider it a system.</p><p>The best illustration of this idea provides in <strong>Figure 1: General types of Engineered System of Interest (SoI) </strong>(‘Types of Systems — SEBoK,’ 2022). It introduces the following concepts:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*fsLE49QRE1KvIa4VtwU90g.png" /></figure><h3>The first layer</h3><p><strong>Technology system of interest</strong>. That system contains the core functionality engineers build, for example, a working prototype of Fortnite in the development environment. Or an operational breadboard for a new industrial computer. Sometimes engineers can call that a product because it results from a technology development project, but in most cases, they use the term “system.” Elements of the development plan on this level are technical capabilities and system requirements.</p><p>Then comes the <strong>“product system of interest.”</strong> This usually goes under the term “product” in systems engineering communities. In the Russian language, there is a very exact term “изделие.” COTS software release, industrial computer, commercial plane, or phone. It has a technology SoI at its core, with a product interface, product control, and related technology. For example, U.I., license, and installation package for software. Or industrial design, warranty, and wireless connectivity for computers. Or Android launcher and shell, platform app store, and branded case design for a phone. Or some particular plane configuration for a transportation company, with several service and warranty contracts and a technological package. The usual approach to distributing roles and responsibilities is that R&amp;D develops technology SoI, and design, marketing, and product organizations are responsible for a product SoI. We plan to release and validate product features as product specifications’ elements on this level.</p><p>But not all products SoI are usable by themselves. Sometimes you need to integrate a <strong>complex solution</strong> from them. Xbox + controller + TV-set + phone + Fortnite. Or CAD/CAM/CAE working station. Or Obsidian + Logseq + Cobian Reflector + Zotero + Office365 + Grammarly setup for a Zettelkasten. It is very similar to the previous one in an engineering sense. In the case of the automobile, we get a highly complex solution already collected together and don’t need to tweak the car in a garage.</p><p>But sometimes, we need to perform such system integration on our own. It could take the form of constructing some design from a T.V. set, self-made antenna, and amplifier like at some point we did in the 80s in the rural areas of Russia (and you needed to know radio technics to operate a T.V. properly). Or we are applying much effort to build a properly working personal computer from components and O.S. distributive in the middle 90s. For that, I will write a separate piece on “integration on the customer side” as it is a fascinating phenomenon of technological advances.</p><p>What’s important to understand here is that we can discuss this system level as a technical package or “product infrastructure.” Innovation happens not in one type of system or product but in some “technological packages.” For example, during the Great Discoveries making open sea navigation devices was a very profitable business, and they improved a lot, but only because of enormous progress in shipbuilding, crew education, cartography, and other parts of the technological package of trans-Atlantic transportation. The same happened with software, the progress of which heavily depended on advances in hardware, network equipment, applied mathematics breakthroughs, and venture funding technologies.</p><p>You cannot ignore and unfollow the evolution of the technological package your product is part of because otherwise, you can find yourself on the losing side of a <a href="https://en.wikipedia.org/wiki/Format_war">format war</a>. I leave alone the current boom of generative A.I. and LLMs, which again results from graphical chips, cloud computing, transformer models, and a lot of manual labor of labelers. What’s important here is that your product is only a part of such “ecosystem” or “infrastructure” or “technology” and shares the proliferation success of the overall solution. And this end-use solution, customers deploy, integrate, and build on their own at the place of operation, fulfilling the role of system integrators. So it would help to have a monitoring and response plan on this level.</p><p>Because of the <a href="https://www.theatlantic.com/technology/archive/2021/06/the-internet-is-a-collective-hallucination/619320/">deterioration of Internet infrastructure</a>, personal search engines need to store referent sources locally or in a cloud or server and not rely on websites as information sources anymore. A phone development team can watch and adapt to Android releases. Streaming companies change the content format following the development of large screens, and game-developing companies respond to the availability of more powerful graphical chips. IKEA adapted the furniture form factor to the typical flat layout. The operational environment matters; sometimes, this environment is part of your innovation wave.</p><p>Not every time this technology package is so apparent. Suppose you are developing a personal productivity tool. In that case, you will track new Notion features release, changes in data exchange APIs for data storage cloud solutions or Rust libraries, and follow the change log of Transformer models: an introduction and catalog — 2023 Edition. If you know your technology package well, small changes (like tweaking API or config) sometimes enable new product features, and you can make a difference in product offers.</p><p>As we see from the cases above, this is not some goal-oriented planning. These are not roadmaps. It is response planning, part of risk and opportunities management, or trust architecture development. The logic of such planning differs significantly from road mapping, so we cannot mix it up without breaking the classification scheme of the plan. Many project and product managers forget that any plan is some taxonomy, a cohesive set of types, and a breakdown containing the same objects (tasks, resources, risks, organizations, features, functions, or locations). And suppose you violate this structure too much. In that case, disorganization will happen (Star, 1995; Bowker and Star, 2000; Garfield, 2023).</p><p>The feature-based and technology package response plans above are perfectly covered by the product line engineering discipline (<em>Product Line Engineering: Are You Missing a Piece in Your Digital Engineering Puzzle?</em>, 2023) and <a href="https://sebokwiki.org/wiki/Capability_Engineering">Capability Engineering — SEBoK (sebokwiki.org)</a> and their planning techniques. A prevalent implementation of this approach for software development is the SAFe framework. This is what ProductPlan in the report called “output/features roadmaps.”</p><p>We see now that this is a required planning level. It’s an essential engineering job. Engineers need product specs because they cannot communicate without them (Madden and Stewart, 2014). And Cagan mentions it explicitly, saying it is the engineering manager’s and CTO’s responsibility to develop a plan at this level. In a perfect world, it is not a job for a product manager (Cagan, 2019). So, there is another planning level and another discipline responsible for it.</p><h3>The second layer</h3><p>The first-level systems provide capabilities that can be packed as products and sold. Planes, cars, phones, and laptops are good examples. Pfizer’s COVID vaccine is a beautiful example of how to win a competition by having the best capability on the market when the product itself (shot) is more or less the same. The business model for product SoI is feature-based, hence output-based roadmaps. There is a feature pack (car options, for example), and each feature gives the user some advantages and delivers benefits. It is a classic FAB sales and marketing technique from the 40s.</p><p>But let’s consider the evolution of technological business models that started in the 2000s. We began to use capabilities to slice them and provide cloud services or merge them to indicate weather and obtain GPS position in smartwatches. We also can build an enormous set of different capabilities, bundle them together, and sell sophisticated tiered access to them as Microsoft does with Office365. Or we make a massive capability but sell only a tiny interface module, as Uber or Alibaba does.</p><p>We must not be mistaken — though we need capabilities for our products to run, it doesn’t mean we should build and operate them. Let’s take the famous case of “the blue box” (<em>Steve Jobs — The Lost Interview</em>, 2013). It was a small device that mimicked the service codes of a telephone network and got you free-of-charge calls. The extensive telecommunication system has provided a proper function, and hackers were selling the product that just abused it, providing unauthorized access to the capability. That brings us to <strong>an essential concept of a service</strong> as “…a mechanism to enable access to one or more capabilities, where the access is provided using a prescribed interface and is exercised consistent with constraints and policies as specified by the service description.” (<em>Reference Ontology for Semantic Service Oriented Architectures v1.0</em>, 2008).</p><p>And many products are exactly like this — Uber app (and service regional operator) is a mechanism to enable access to the transportation system; your mobile bank app (together with the frauds department and customer support) is a mechanism to facilitate access to the banking system; and chatGPT is a mechanism to enable access to a massive generative A.I. system with a significant social component.</p><p>We sell a product (Haines, 2021), and in the case of service SoI, its boundaries do not coincide with capabilities boundaries. Services live on top of capabilities — taxi companies, banks, and LLMs have already existed with another service layer. We can rebuild the service layer separately with very few tweaks of the core capability to make a successful product from a failing system. Here are a couple of cases.</p><p>One research university deployed a comprehensive Wi-Fi campus network. Though technologically, it was a very successful project (capability was ok), the dean was unhappy. The network was expensive, and students overloaded its segments when large groups visited some particular site simultaneously. The CTO asked for additional equipment to address such complaints, but they could not be justified. The project was already too expensive. So the dean appointed the product manager to fix the problem. And with just a few small changes in user experience and network reconfiguration, the PM introduced paid tariffs that allowed prime access for a small charge and other product modifications. With its budget from these payments, the project survived and thrived.</p><p>The online trader’s fulfillment center has had difficulties measuring the effectiveness of pick-up procedures. The PM came up with the idea of using smartwatches to monitor the movements of warehouse and delivery operators. It took only small API changes to collect the data from bracelets. Again, there was no capability change, but it ultimately resulted in another system function (a smartwatch is an indicator and autonomous sensor module, speaking of system functions) and service. And it was a much more successful product than the classical warehouse monitoring solutions industry used.</p><p>Again, the above figure explains the systems engineering concept of service very well. The Fortnite service, for example, includes subscriptions, payments, marketing, tournaments, regional servers, a map editor, customer support, updates, match-making, etc. We don’t need product line or capabilities engineering to plan these activities; we have service science and service engineering. Technical capabilities of Fortnite technological core are well outside of game service, and service development requires separate planning, with service elements, not capabilities, as objects of planning. A couple of years ago, I heard a story about how Grundfos turned their pumps into service SoI with subscriptions, and my guess is it did not take very much capabilities engineering. The same happened with service contracts for plane engines and car subscriptions.</p><p>Services isolate you from capabilities and give you an abstracted technology interface. And you have to plan their development and deployment separately from the product SoI that delivers capabilities. Look at the OASIS standard I cited above; it gives an excellent conceptualization of service systems (and for the categories of service development planning) concisely.</p><h3>The third level</h3><p>And, finally, here goes the “layer of products” as product managers from Meta or SVPG write about them (Cagan, 2022; Ganti, 2023) — products as a set of practices for making money. When we label “product,” we just want to say that we understand how we make money. Or, making the definition more strict, <em>“a product is a construct that reduces uncertainty in practices of making money.”</em></p><p>Technologically a product can be minimal, like the Uber app. It is a functionally insignificant part of the transportation system of systems that provides functional capability and services. The same applies to the YouTube app and YouTube system and services on top of it, or chatGPT, or Carrefour online store, which is a small visible part of the enormous retail chain. Product managers discover such parts and monetize them. Users do not consider the value of some hidden capabilities, and you cannot charge them in service models.</p><p>Critical remark regarding the cost of products — in systems engineering planning, we look at a project with the eyes of a very qualified investor and operator who wants to get into the details of everything. System-of-interest, enabling systems, the operational environment, and all lifecycle costs must be calculated and justified with economic effects and benefits. But when people were buying Amazon AWS, they didn’t want to think about all that sunk costs of data centers and the necessity to return these investments for Amazon, so carefully designed product-as-practices provided precisely that. They limit concepts only to your operational domain, so you perceive it as “jobs-to-be-done,” and not system functions that can have emergent behavior and adverse effects.</p><p><strong>So, when you consider Fortnite products, you have all three levels.</strong> You pay for the product SoI, which is the application itself, you can pay for the service Xbox Game Pass to access the game, and finally, you can pay for Fortnite Crew and other perks and bonuses. All three levels work cohesively to hold you in the environment where you spend money. Seasonal campaigns are free of charge, but they reduce churn, and some gamers who stay will pay for a Fortnite Pass and Crew. The same goes with YouTube — you can pay for a premium subscription, or businesses pay for ads you watch. It’s all the product manager’s job, and they have their disciplines, which are beautifully described in (Steinhardt, 2017). And when I hear “outcome-based roadmaps,” I understand that people want this third type of “product.” And when you think about products like systems engineer, how can you sum up technological SoI and product-as-practices and demand they must produce 100% together? They cannot.</p><h3>Bibliography</h3><p><em>Apollo-Protocol/4d-activity-editor</em> <em>GitHub</em>. Available at: <a href="https://github.com/Apollo-Protocol/4d-activity-editor/discussions">https://github.com/Apollo-Protocol/4d-activity-editor/discussions</a> (Accessed: 18 May 2023).</p><p>Bowker, G.C. and Star, S.L. (2000) <em>Sorting things out: Classification and its consequences</em>. MIT Press.</p><p>Cagan, M. (2019) ‘Product vs. Feature Teams,’ <em>Silicon Valley Product Group</em>, 29 August. Available at: <a href="https://www.svpg.com/product-vs-feature-teams/">https://www.svpg.com/product-vs-feature-teams/</a> (Accessed: 26 April 2022).</p><p>Cagan, M. (2022) <em>The Nature of Product</em>, <em>Silicon Valley Product Group</em>. Available at: <a href="https://www.svpg.com/the-nature-of-product/">https://www.svpg.com/the-nature-of-product/</a> (Accessed: 25 April 2022).</p><p>Ganti, A. (2023) ‘How Meta uses analytics to assess Product Market Fit’, <em>Medium</em>, 21 February. Available at: <a href="https://aniganti.medium.com/how-meta-uses-analytics-to-assess-product-market-fit-c9262add2790">https://aniganti.medium.com/how-meta-uses-analytics-to-assess-product-market-fit-c9262add2790</a> (Accessed: 22 May 2023).</p><p>Garfield, S. (2023) ‘Classification Process and Taxonomy’, <em>Medium</em>, 23 March. Available at: <a href="https://stangarfield.medium.com/classification-process-and-taxonomy-dbefa98bb396">https://stangarfield.medium.com/classification-process-and-taxonomy-dbefa98bb396</a> (Accessed: 27 March 2023).</p><p>Haines, S. (2021) <em>The Product Manager’s Desk Reference, Third Edition</em>. 3rd edition. New York: McGraw Hill.</p><p>Kun, J. (2016) ‘Habits of highly mathematical people’, <em>Medium</em>, 5 August. Available at: <a href="https://medium.com/@jeremyjkun/habits-of-highly-mathematical-people-b719df12d15e">https://medium.com/@jeremyjkun/habits-of-highly-mathematical-people-b719df12d15e</a> (Accessed: 25 October 2022).</p><p>Madden, J. and Stewart, R. (2014) ‘One Hundred Rules for NASA Project Managers’, p. 13.</p><p>Oehmen J. et al. (2012) <em>The guide to lean enablers for managing engineering programs</em>. Joint MIT‐PMI‐INCOSE Community of Practice on Lean in Program Management.</p><p>Partridge, C. <em>et al.</em> (2020) ‘A survey of Top-Level Ontologies To inform the ontological choices for a Foundation Data Model Version 1’. Construction Innovation Hub. Available at: <a href="https://www.cdbb.cam.ac.uk/files/a_survey_of_top-level_ontologies_lowres.pdf.">https://www.cdbb.cam.ac.uk/files/a_survey_of_top-level_ontologies_lowres.pdf.</a></p><p>Partridge, C. (2021) ‘A 4-Dimensionalist Top Level Ontology (TLO): Mereotopology and Space-Time’, in. <em>Newton Gateway to Mathematics</em>, INI Seminar Room 1. Available at: <a href="https://gateway.newton.ac.uk/presentation/2021-04-22/29946">https://gateway.newton.ac.uk/presentation/2021-04-22/29946</a> (Accessed: 18 November 2022).</p><p><em>Product Line Engineering: Are You Missing a Piece in Your Digital Engineering Puzzle?</em> (2023). (Calling All Systems). Available at: <a href="https://www.youtube.com/watch?v=uQy6RBfguhw">https://www.youtube.com/watch?v=uQy6RBfguhw</a> (Accessed: 29 May 2023).</p><p><em>Reference Ontology for Semantic Service Oriented Architectures v1.0</em> (2008). Available at: <a href="http://docs.oasis-open.org/semantic-ex/ro-soa/v1.0/pr01/see-rosoa-v1.0-pr01.html#_Toc214854765">http://docs.oasis-open.org/semantic-ex/ro-soa/v1.0/pr01/see-rosoa-v1.0-pr01.html#_Toc214854765</a> (Accessed: 26 February 2023).</p><p>Star, S.L. (1995) ‘The politics of formal representations: Wizards, gurus, and organizational complexity’, <em>Ecologies of knowledge: Work and politics in science and technology</em>, 88, p. 118.</p><p>Steinhardt, G. (2017) <em>The Product Manager’s Toolkit®</em>. Cham: Springer International Publishing (Management for Professionals). Available at: <a href="https://doi.org/10.1007/978-3-319-49998-7.">https://doi.org/10.1007/978-3-319-49998-7.</a></p><p><em>Steve Jobs — The Lost Interview</em> (2013). Available at: <a href="https://www.youtube.com/watch?v=TRZAJY23xio">https://www.youtube.com/watch?v=TRZAJY23xio</a> (Accessed: 26 April 2022).</p><p>‘The 2023 State of Product Management Report’ (2023). ProductPlan. Available at: <a href="https://assets.productplan.com/content/The-State-of-Product-Management-Annual-Report-2023.pdf.">https://assets.productplan.com/content/The-State-of-Product-Management-Annual-Report-2023.pdf.</a></p><p>‘Types of Systems — SEBoK’ (2022). Available at: <a href="https://sebokwiki.org/wiki/Types_of_Systems">https://sebokwiki.org/wiki/Types_of_Systems</a> (Accessed: 28 May 2023).</p><p>West, M. (2011) <em>Developing high quality data models</em>. Elsevier.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d49b4c00e09a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/three-layers-of-products-conceptualization-framework-d49b4c00e09a">Three layers of “products”: conceptualization framework</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[When Shu-Ha-Ri flips to Ri-Ha-Shu]]></title>
            <link>https://alex-turkhanov.medium.com/when-shu-ha-ri-flips-to-ri-ha-shu-821424a21b1?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/821424a21b1</guid>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Sat, 20 May 2023 06:46:30 GMT</pubDate>
            <atom:updated>2023-05-20T07:05:36.083Z</atom:updated>
            <content:encoded><![CDATA[<p><em>some occasional thoughts about exams and trials</em></p><p>I have been passing a lot of tests and certifications lately. That is what comes with moving to another country. Yesterday I was taking a driving exam in Spain (my country license is not valid here, and you need to obtain a new one from the beginning). That means you must follow stringent traffic rules when everyone else doesn’t, which is a double challenge. Strict rules mean you get half the limit of your elimination penalty for “excessively waiting” for 2–3 seconds on the busy roundabout entry if a car is behind you. Road users do not follow them here.</p><p>When you go through the list of mistakes for which you get penalty points, you will see two types of rules embedded in them. Actual safety rules and “effective use of road and vehicle” rules. In other words, there is a mandatory style of driving they check during the exam — keeping the fluidity of traffic, keeping the road _actually_safe- and also _perceived_safe_ by the police, older citizens, parents, and so on. Such rules altogether I call “style.”</p><p>And, of course, this “style” is unnatural. Actual driving practices are quite different; it is a utopian (in James Scott’s definitions) perspective of the state of how things should be. And so, to pass the exam, you need to unlearn your actual driving skills and learn this new style. You must abandon your Ri for this Shu only to recover it later.</p><p>Somehow it reminds me of two things:</p><p>1) The so-called “state-of-the-art knowledge” that some educational institutions offer. “We have completely updated the discipline, and the course is new, so you need to retake your certification because your knowledge is obsolete now.” Well, I was a master and did well, but now I am a novice again, and somehow I am failing at my work just because some professor discovered a few books and articles for himself? I am not sure it works that way; sorry, I pass.</p><p>2) The HR recruiter interview, when you need to “explain in simple terms” what you did and why it is so great. It’s like trying to rewrite a very technical article for a newspaper. So much sense gets lost that it becomes a nuisance.</p><p>It is very hard — to go down from Ri to Shu when it is Shu who judges Ri. The yellow belt assesses the black belt; does he do well or not? Stupid, right?</p><p>But there is a practical way out. First, turn this exam into sociology research — find what theory explains the successful behavior; discover and test variables that prove your assumptions were correct; learn the story and tell it to your friends (in Spanish). Second, learn the new style and its patterns. In the end, this is your professional challenge.</p><p><strong>Update.</strong> Exam introduces you to some imaginary, idealized world where the exam concepts comprise a cohesive domain model. The real world is dirty and nothing like this imaginary one. This notion stands for any exam. What you, as a professional, must do is imagine this world when you pass the exam.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=821424a21b1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Sack (1950), LLMs, and humans in the loop]]></title>
            <link>https://alex-turkhanov.medium.com/the-sack-1950-llms-and-humans-in-the-loop-d8497751407a?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/d8497751407a</guid>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Fri, 19 May 2023 07:52:54 GMT</pubDate>
            <atom:updated>2023-05-19T07:52:54.957Z</atom:updated>
            <content:encoded><![CDATA[<p>Do you remember The Sack, a 1950 science fiction story by William Morrison? I recently re-read it, and the story describes one possibility we can fail as communities of researchers. We can lose a lot of side-kick opportunities to discover something along the way to the main goal if we immediately find a convincing answer we initially imposed. And our research protocol relies on such sudden discoveries so heavily that when adapted to the practical task of software development, we came up with Agile methods and opportunistic/evolutionary life cycles.</p><p>The same degradation of search skills can happen if LLMs can provide a good answer for a complicated answer that now would require some research, trial, and error, but in exchange, extend our knowledge in the initial direction and the others we did not know before. Such scenarios already happened before; I can bring up the degradation of flight skills of pilots that occurred in Russian civil aviation after fleet operators’ management started to rely heavily on flight automation to exclude the “human factor.” There are other examples of how technology destroys human abilities to navigate and orient.</p><p>Technology does not include people, but systems that use this technology do. When we evaluate the pros and cons, we cannot take the people out of the equation; they are part of it. The Sack in Morrison’s story was such a technology, but it was just a small part of a more extensive system that emerged around it. The same can happen with LLMs. They can have strong emergent behavior.</p><p>Both technology optimists (LLMs are for good) and technology skeptics (LLMs do not matter) need to include one crucial factor we as systems engineers do — humans in the loop. LLMs are not just software systems anymore. They have operators. And there’s more — these operators are not adequately trained, so they can become LLM’s actuators — gullible people who can execute Bard’s instructions pretty well (they execute stupid commands from Google Maps after all, don’t they?) and don’t have guardrails.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d8497751407a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Failure of product metaphor in Word]]></title>
            <link>https://alex-turkhanov.medium.com/failure-of-product-metaphor-in-word-d53a3c1a8b10?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/d53a3c1a8b10</guid>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Wed, 10 May 2023 08:03:51 GMT</pubDate>
            <atom:updated>2023-05-10T08:03:51.398Z</atom:updated>
            <content:encoded><![CDATA[<p>What has surprised me for a long time in Office 365 is that with all their top-of-the-shelf product management, their product still misses the critical feature of the document panes.</p><p>For example, I tried to write the last article in Word instead of the usual flow in Obsidian or Logseq, and I couldn’t. When you write an article, you have a research focus (thesis, hypotheses, questions, method) written in the front and arguments across the text. And when you write, you should look at them not to lose focus.</p><p>It is straightforward in Obsidian or Logseq but challenging when doing the same in Word. It just doesn’t have the functionality of panes or sidebar or clipboard. They still use the same old text processor metaphor without changing it.</p><blockquote>It’s the same problem with YouTube, for example. As a creator, I always fought with interpretations of calls to action and content efficiency through likes. They are missing the timestamp feature. That would be a considerable improvement — I would know precisely when they reacted. Was my CtA compelling? What part of the content engages them best?</blockquote><p>I wrote about the idea of product metaphor in one of my articles and still think <a href="https://www.linkedin.com/feed/hashtag/?keywords=productmanagement&amp;highlightedUpdateUrns=urn%3Ali%3Aactivity%3A7061969338513133568">#productmanagement</a> and <a href="https://www.linkedin.com/feed/hashtag/?keywords=systemsengineering&amp;highlightedUpdateUrns=urn%3Ali%3Aactivity%3A7061969338513133568">#systemsengineering</a> will benefit from burrowing science and technology studies concepts and adapting them for their frameworks. For example, in the last 5th edition of the SE Handbook, INCOSE emphasizes (again, with the new reference) that products = systems. Well, they are not if you look from the microsociology and STS perspective, and this was the thesis from the article I started the story with.</p><p>Anyway, it is a whole new world of product management for me to research from SE viewpoints.</p><p><a href="https://lnkd.in/d68Dw2iK">https://lnkd.in/d68Dw2iK</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d53a3c1a8b10" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The flawed metaphor of parenting]]></title>
            <link>https://alex-turkhanov.medium.com/the-flawed-metaphor-of-parenting-7a138a2509f9?source=rss-4953e9190086------2</link>
            <guid isPermaLink="false">https://medium.com/p/7a138a2509f9</guid>
            <dc:creator><![CDATA[Alexander Turkhanov]]></dc:creator>
            <pubDate>Fri, 14 Apr 2023 08:34:18 GMT</pubDate>
            <atom:updated>2023-04-14T08:34:18.270Z</atom:updated>
            <content:encoded><![CDATA[<p>Not very long ago, I ran into a story where the author explained “why being a mom is a CEO job,” with many supportive comments and reactions under the publication. While this could be an attractive narrative, it steers us from thinking straight. I used such “family metaphors” myself, and I was wrong; being a parent is not comparable to a CEO job. Here are the reasons.</p><p>1. There is no competition. Really, if you are failing as a parent, it is not like other families will come and take away your children, your income, and your household. At least not in many societies, and such cases, it’s more like to become a subject of regulatory control, not competition. You don’t compete on the market to be where you regard your family. At least not at the same level as a CEO running a company does.</p><p>2. There is no IT and infrastructure to manage (really). Even small businesses run a handful of IT systems and are very concerned with the solutions and how they use them. It affects operations a lot in every significant aspect — from customer acquisition to product and service delivery. You must be good at setting up and running IT infrastructure if you want to get some money and some sleep. When you are a parent, it does not matter very much. You can use a notebook and a pencil and still be good at your job.</p><p>3. There are no clear goals, metrics, compliance, or governance for parents. Maybe, in some more traditional societies, there can be some “family board meeting,” but in no way is it so obligatory as the board of directors for the CEO. You don’t have to comply with hundreds of laws, the charter, meet the targets, or otherwise explain why you are doing what you are doing. So much more manageable, with freedom and control over your life. Nothing compared to a CEO life.</p><p>Metaphors are very attractive, but we must be careful when using them. We don’t need to hallucinate now; we have a chatGPT for that. If we bring some metaphor into the play, let’s think clearly with data. If a CEO job is like parenting, what percentage do we put on the similarity? How do we calculate this number? Running what families are more like a CEO job and why? And if there are no at least vague explanations to these questions, can we not bring the potentially flawed metaphor into the discussion?</p><p>When writing this piece, I saw <a href="https://www.linkedin.com/feed/update/urn:li:activity:7052236517045334016/">the post</a> where SW development planning had been compared to orienting with GPS. Below is my comment on it.</p><p>This metaphor is broken in an essential aspect. Though the project or product team navigates the space seeking success and the ship’s crew, these spaces are very different. Let’s see.</p><p>When we use GPS to understand our location and where we need to be, we use infrastructure facts. There is enormous scientific and engineering machinery proving that I am currently at these coordinates — global atomic clock, Internet, GPS constellation, GPS-certified device, geography and trigonometry, navigational mechanics, Boolean logic, metrology, and who knows what else.</p><p>When I say I am at some coordinates at some point in time, I rely on this infrastructure, and if someone wants to prove me wrong, they need to prove that the whole infrastructure is flawed. And that’s extremely difficult. The fact that I am where I am at the exact point of time using the GPS data is solid and infrastructural. And with SW development, we don’t have the universal understanding of Euclidian space and global infrastructure that, in a standard way by a standard interface, will tell every project team where they are precisely right now and what maneuvers they need to perform to be at the destination point.</p><p>It’s much like navigation in the Great Discoveries era, where Cristóbal Colón was entirely sure until his death that he found East India, and such minor facts that the star map was utterly different could not change his mind. The location without coordinates is a social fact, not an infrastructural one.</p><p>That changes the discussion because, in a debate about where we are and where we need to go, we cannot use math, science, or engineering measurements. We seek social consensus. And that’s where the metaphor is broken.</p><p>Fix your metaphor before using it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7a138a2509f9" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>