{"id":138649,"date":"2026-03-03T11:10:39","date_gmt":"2026-03-03T11:10:39","guid":{"rendered":"https:\/\/sii.pl\/?p=138649"},"modified":"2026-03-30T10:57:53","modified_gmt":"2026-03-30T10:57:53","slug":"ai-architect-projects-challenges-and-growth","status":"publish","type":"post","link":"https:\/\/sii.pl\/en\/news-feed\/ai-architect-projects-challenges-and-growth\/","title":{"rendered":"AI Architect&#8217;s role: Projects,\u00a0challenges\u00a0and growth"},"content":{"rendered":"<div class=\"wp-block-sii-nsw-container container container-8349c92d-69c5-4587-9be4-7b93116698ed\"><style type=\"text\/css\">.container-8349c92d-69c5-4587-9be4-7b93116698ed {  }\n                         @media screen and (max-width: 991px) { .container-8349c92d-69c5-4587-9be4-7b93116698ed {  } }<\/style><h3 class=\"wp-block-heading\"><strong>With the growing role of AI\/ML in business and increasing interest in leveraging AI, organizations are increasingly looking for professionals who can combine technology, data, and business goals into a coherent and effective AI system architecture, while also being able to code, implement solutions in practice, and ensure their optimization. <\/strong><\/h3>\n\n<p>In this interview, we talk to Alan, an AI\/ML Architect at Sii Warsaw, who has been designing and implementing AI\/ML, GenAI, and data-driven systems in production environments for years. If this is the direction you would like to grow in, check out the current&nbsp;<a href=\"https:\/\/sii.pl\/en\/job-ads\/\" target=\"_blank\" rel=\"noopener\" title=\"\">job&nbsp;offers&nbsp;at&nbsp;Sii<\/a>.&nbsp;<\/p>\n\n<h2 class=\"wp-block-heading\">Who&nbsp;is&nbsp;an&nbsp;AI\/ML Architect?&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">How&nbsp;would&nbsp;you&nbsp;describe&nbsp;your&nbsp;role in a&nbsp;few&nbsp;sentences?&nbsp;<\/h3>\n\n<p>The AI\/ML Architect role goes far beyond simply designing system architecture and using AI-based tools. Depending on the project, it combines the competencies of a tech lead, software developer, DevOps engineer, and sometimes even a data analyst. I see this role in line with Gartner\u2019s definition, where an IT architect&nbsp;is responsible for&nbsp;designing solutions that connect business needs with technological capabilities. <\/p>\n\n<p>An AI\/ML Architect looks at the system holistically, drawing on experience in design and implementation, making technological decisions, and taking responsibility for their long-term consequences. Extensive technical knowledge is essential. Balancing code, data, architecture, and business is also crucial. Even the best AI solution, without production deployment,&nbsp;remains&nbsp;only&nbsp;proof&nbsp;of concept.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">What&nbsp;does&nbsp;your&nbsp;typical&nbsp;workday&nbsp;look&nbsp;like?&nbsp;What&nbsp;can&nbsp;be&nbsp;frustrating&nbsp;in&nbsp;this&nbsp;role?&nbsp;<\/h3>\n\n<p>There is no single \u201ctypical\u201d&nbsp;day, because&nbsp;my role combines architecture, software development, data work, and collaboration with business stakeholders. Part of the day involves project meetings and consultations with ML, data engineering, or DevOps teams, while another part focuses on designing AI\/ML system architecture and conducting technical reviews. <\/p>\n\n<p>In practice, a lot of time is spent making decisions that have long-term consequences for the entire software ecosystem &#8211; from technology selection and data processing approaches to AI\/ML production deployment strategies.&nbsp;<\/p>\n\n<p>What&nbsp;frustrates&nbsp;me? The&nbsp;number&nbsp;of&nbsp;meetings&nbsp;and&nbsp;sometimes&nbsp;rigid&nbsp;processes&nbsp;&#8211;&nbsp;not&nbsp;every&nbsp;project&nbsp;has&nbsp;to&nbsp;look&nbsp;the same.&nbsp;However, we&nbsp;actively&nbsp;work&nbsp;on&nbsp;improving&nbsp;that.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">What&nbsp;competencies&nbsp;are&nbsp;key&nbsp;for&nbsp;this&nbsp;AI role?&nbsp;<\/h3>\n\n<p>Adaptability and familiarity with multiple technologies are essential, as they allow effective communication with other teams and architects when developing shared solutions. <\/p>\n\n<p>A few pieces of advice for aspiring architects: learn broadly, not only deeply&nbsp;&#8211;&nbsp;explore different technologies. Understand the full project lifecycle, from data, through the model, to deployment and monitoring. Learn to read other people\u2019s code&nbsp;&#8211;&nbsp;you will read more than you write. Look for projects where you can take end-to-end responsibility.&nbsp;Don\u2019t&nbsp;be afraid to&nbsp;say,&nbsp;\u201cI don\u2019t know.\u201d Make mistakes but learn from them.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">What&nbsp;non-technical&nbsp;skills&nbsp;are&nbsp;important&nbsp;in&nbsp;this&nbsp;role?&nbsp;<\/h3>\n\n<p>Over the years, I have had to develop a wide range of soft skills. Work organization and time management. Leading constructive discussions. Negotiation and handling conflicts. The ability to work with people from different fields and cultures and to understand their perspectives. Communication is especially important \u2013 the ability to translate technical matters into business language and vice versa, so that both sides truly understand each other. Convincing others of your idea is not manipulation \u2013 it is explaining and jointly discovering solutions. Active listening, empathy, and the ability to adapt to different communication styles within a team are the foundations of building trust. <\/p>\n\n<p>You can be a brilliant programmer, but if you cannot talk to people and present your ideas, it will be difficult to persuade others to follow you. Non-technical skills are not an addition \u2013 they are at the core of an architect\u2019s craft. Without them, even the best technical decisions may never materialize into real solutions. This is what distinguishes architects who truly influence an organization from those who remain at the proposal stage.<\/p>\n\n<h3 class=\"wp-block-heading\">Which personality traits help, and which&nbsp;complicates&nbsp;being a good architect?&nbsp;<\/h3>\n\n<p>Curiosity,&nbsp;patience&nbsp;(which&nbsp;I&nbsp;am&nbsp;still&nbsp;working&nbsp;on),&nbsp;empathy, and&nbsp;listening&nbsp;skills.&nbsp;<\/p>\n\n<p>What&nbsp;gets&nbsp;in the&nbsp;way?&nbsp;Perfectionism&nbsp;&#8211;&nbsp;you&nbsp;cannot&nbsp;do&nbsp;everything&nbsp;perfectly, and&nbsp;you&nbsp;must&nbsp;accept&nbsp;that&nbsp;compromises&nbsp;will&nbsp;always&nbsp;appear.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">How&nbsp;does&nbsp;an&nbsp;AI\/ML Architect&nbsp;differ&nbsp;from a Data&nbsp;Scientist&nbsp;and&nbsp;an&nbsp;AI&nbsp;Engineer?&nbsp;<\/h3>\n\n<p>A Data Scientist and an AI Engineer focus on specific areas such as data analysis, model training, or building ML pipelines. An AI\/ML Architect connects these perspectives and designs the entire AI ecosystem \u2013 from data and models to integration with business systems. The key skill is \u201cconnecting the dots\u201d between AI\/ML systems and understanding the impact of a single component on the entire software ecosystem. <\/p>\n\n<p>At the beginning of a project, I ask questions that help define the problem and set its boundaries: What is the time horizon? Will the solution work when the load increases several times over? What happens when the data changes? How will we monitor it? This is just the beginning, but it helps us understand what we are dealing with and which direction to take.<\/p>\n\n<h2 class=\"wp-block-heading\">Designing AI\/ML&nbsp;systems&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">What does the AI\/ML architecture design process look like?&nbsp;<\/h3>\n\n<p>I always start with the data. Where does it come from? What is its quality? How is it stored? What does its actual flow within the organization look like? <\/p>\n\n<p>The key is understanding not how procedures say something should work, but <strong>how it really works<\/strong>. Often, it turns out that data partially functions outside systems \u2013 in Excel files, documents, or manual processes. This has a huge impact on whether an AI\/ML project has any chance of success at all. <\/p>\n\n<p>The next step is to understand the business problem. What exactly do we want to achieve? How will we measure success? Do we really need a language model, or would a simpler solution be enough? Only later do architectural and technological decisions come into play. Every project is different, so the approach must be adapted to the situation. Any AI-based project requires not only working with models, but also solid system engineering.<\/p>\n\n<h3 class=\"wp-block-heading\">What architectural decisions are key&nbsp;to&nbsp;AI\/ML projects?&nbsp;<\/h3>\n\n<p>In AI\/ML projects, we make many decisions with long-term consequences. We need to answer questions: Should we build our own solution or use an existing one? How should data flow through the process, and where will it be stored? How much complexity should we add now, and how much should we leave for later? Should we process data in real time or in batches? <\/p>\n\n<p>There is one trap I am particularly cautious about: choosing the core technology too early. We often stay with chosen technology for a long time. A decision made too quickly can follow the entire project and start causing problems. With every idea, I try to consider: Can we reuse this AI-based solution elsewhere? Are we already doing something similar in another system? How can we limit work to avoid duplicating effort? I try to see connections between systems that others might not notice.<\/p>\n\n<h2 class=\"wp-block-heading\">AI in&nbsp;business&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">Why are companies increasingly investing in AI\/ML?&nbsp;<\/h3>\n\n<p>Companies have been implementing AI\/ML solutions for years. However, the popularization of GenAI and models such as ChatGPT has made AI more understandable and accessible to business stakeholders. Technology has become tangible to everyone, but at the same time there is a tendency to use it where it is not always justified. The Architect\u2019s role is also to clearly state when AI\/ML makes business sense and when a classic approach would be more effective. Well-designed AI\/ML solutions can genuinely increase ROI and automate processes.<\/p>\n\n<h3 class=\"wp-block-heading\">How do you balance business requirements with technical&nbsp;limitations?&nbsp;<\/h3>\n\n<p>I try to look for solutions instead of saying something cannot be done. Sometimes implementation requires more time, process changes, or a different approach than initially assumed. If something is impossible at a given moment, I propose an alternative, for example, starting data collection and returning to the topic in a few months. When communicating with business stakeholders, I avoid technical jargon. I use analogies and examples that show the scale of the problem. The goal is to share an understanding of constraints, not proving someone wrong.<\/p>\n\n<h3 class=\"wp-block-heading\">What risks most often appear in AI projects? Can you&nbsp;give me&nbsp;an example?&nbsp;<\/h3>\n\n<p>AI projects involve many challenges, and identifying and managing them requires a mature approach. I will mention several areas I most often observe: gaps in system architecture, underinvestment in infrastructure, and insufficient capacity planning. <\/p>\n\n<p>However, data-related issues remain the most common, widespread, and underestimated risk in the entire AI industry \u2013 an almost universal challenge affecting most projects. Companies collect data for years, but without consistent rules. Lack of information about who changed the data and when, loss of important information during migrations, or data that represents nothing and is full of noise or logical inconsistencies. <\/p>\n\n<p>Additional complexity comes from scale and data dispersion. We work with databases containing hundreds of thousands of records; they come from more than a dozen different systems and applications, each with its own format, encoding, and terminology. There is no information about changes, noise appears, inconsistencies arise, migrations destroy part of history. An example from my practice: the process assumed a single document template for entering data. <\/p>\n\n<p>Over the years, that document evolved, was modified during completion, and alternative versions were created. We ended up with several thousand versions of \u201cthe same\u201d document. We technically have the data, but extracting consistent conclusions from it is very difficult. At the beginning of a project, we check what is in the data, what it looks like, what is missing, and we confront that with the process and the real information flow. Today, GenAI also helps us address some of these issues more easily.<\/p>\n\n<h3 class=\"wp-block-heading\">How do you assess whether an AI solution truly delivers business value?&nbsp;<\/h3>\n\n<p>I look at direct value: time savings, cost reduction, process automation. But value also includes team competency development and improving data quality across the entire ecosystem. One AI project can positively impact many other systems.<\/p>\n\n<h2 class=\"wp-block-heading\">AI Engineer &#8211; what technologies are used?&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">Which technologies are most important in your work?&nbsp;What\u2019s&nbsp;overrated, and&nbsp;what\u2019s&nbsp;underrated?&nbsp;<\/h3>\n\n<p>Key criteria for tool selection: maturity \u2013 active community, market recognition, real production implementations. GenAI has turned this upside down. Previously, it was easy to identify valuable libraries \u2013 hundreds to a few thousand GitHub stars, active development. Now, something new can appear and gain tens of thousands of stars within weeks simply because it\u2019s related to GenAI. Maturity is lacking. New tools emerge rapidly and gain popularity, but stability doesn\u2019t always follow. <\/p>\n\n<p>Overrated? The belief that GenAI will solve every problem. Sometimes a simple algorithm or proper regex is enough. RAG is great for specific applications, not a universal solution. <\/p>\n\n<p>Underrated? ML monitoring and observability, simple heuristics, rule-based systems, and data governance. If using LLMs, proper data preparation and systematic evaluation of model outputs are crucial.<\/p>\n\n<h3 class=\"wp-block-heading\">How does collaboration with other&nbsp;teams&nbsp;work?&nbsp;<\/h3>\n\n<p>AI\/ML projects often involve multiple teams: research, application, backend, MLOps. In our team, we combine these roles, but we also collaborate with integration, infrastructure, front-end, and QA.<\/p>\n\n<h3 class=\"wp-block-heading\">How do you approach&nbsp;MLOps&nbsp;and model monitoring?&nbsp;<\/h3>\n\n<p>Monitoring is fundamental for production models. We collect logs, set error alerts, track latency and endpoint load, and monitor business metrics, user counts, and response quality. This way, if a user reports a problem, we can trace every step. We have automated CI\/CD. As dependent applications grow, deployment planning becomes crucial. Monitoring also enables automated retraining: we can detect data drift and verify whether the model degrades or results no longer meet expectations. Then we know it\u2019s time to retrain.<\/p>\n\n<h2 class=\"wp-block-heading\">Experience and&nbsp;career&nbsp;path&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">What did your path to becoming an AI\/ML Architect look like?&nbsp;<\/h3>\n\n<p>My path wasn\u2019t planned. I started as a C++ developer in telecommunications. I discovered Python by chance through a colleague, fell in love with it, and it quickly became my main tool. Python seems simple, but to unlock its full potential, you must dive deep. <\/p>\n\n<p>Later, I joined an ML project as support because I knew Python and C++. That\u2019s when I discovered data science. I also had a stint in banking with Big Data and Spark, which gave me insights into different data processing approaches. I\u2019ve always been curious about how systems work end-to-end, exploring many perspectives. I got involved, took initiatives, and got the chance to lead an entire project. That\u2019s when I realized this was my path. <\/p>\n\n<p>I have implemented AI solutions in production for telecom and banking clients, and at Sii, several large AI projects and dozens of smaller ones, often based on or reusing previously deployed systems and solutions. I\u2019m not an architect \u201cby definition\u201d; I combine many roles in one. The current market expects this broader perspective.<\/p>\n\n<h3 class=\"wp-block-heading\">Can a developer or data scientist smoothly transition to an architect role?&nbsp;<\/h3>\n\n<p>Yes, but it requires a change of perspective. Both groups often lack the view that AI\/ML&nbsp;isn\u2019t&nbsp;just a model or code. It involves stakeholders, business processes, maintenance, and development. Even the best model, if it cannot be deployed or integrated into a business process,&nbsp;remains&nbsp;just a model. Great code that&nbsp;doesn\u2019t&nbsp;solve a real problem&nbsp;&#8211;&nbsp;or makes it harder&nbsp;&#8211;&nbsp;is not a solution.&nbsp;<\/p>\n\n<h2 class=\"wp-block-heading\">Work culture and development&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">How does&nbsp;Sii&nbsp;support employee development?&nbsp;<\/h3>\n\n<p>Sii&nbsp;supports employee development broadly. We have many internal&nbsp;training courses&nbsp;and opportunities for external courses. Certification is also possible.&nbsp;Sii&nbsp;hosts multiple initiatives&nbsp;&#8211;&nbsp;communities, meetups, and events&nbsp;&#8211;&nbsp;and employees can attend conferences. Importantly,&nbsp;Sii&nbsp;supports not only technical development. Initiatives like sponsoring employees\u2019 passions or volunteering are equally important. There are many development opportunities; it depends on your engagement and motivation.<\/p>\n\n<h3 class=\"wp-block-heading\">How is knowledge&nbsp;exchange&nbsp;between teams?&nbsp;<\/h3>\n\n<p>We have constant contact, discussing approaches, and solutions. For example, all architects meet regularly. Knowledge sharing is not limited to architects or AI. When exploring&nbsp;new technology,&nbsp;it\u2019s&nbsp;worth consulting people with experience,&nbsp;both architects and engineers.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">Is there&nbsp;a place&nbsp;for experimentation?&nbsp;<\/h3>\n\n<p>My passion is&nbsp;new technologies, so I often experiment with new libraries, solutions, and approaches. We have access to the latest models and space to experiment. Deadlines are tight, and work is heavy, but there\u2019s always room to develop and test&nbsp;new solutions. Without this,&nbsp;there\u2019s&nbsp;no innovation, and I focus on innovation. A concept may not&nbsp;immediately&nbsp;be production-ready, but some ideas can be used in current solutions or in developing new ones. Experimentation is therefore an essential part of our work.&nbsp;<\/p>\n\n<h2 class=\"wp-block-heading\">AI Developer &#8211; salary and motivation&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">What&nbsp;determines&nbsp;an AI\/ML Architect\u2019s salary?&nbsp;<\/h3>\n\n<p>Many factors &#8211; experience, career history, completed projects. Sometimes also the company, its financial capacity, and project complexity. Much depends on the individual. One person may prefer frequent changes and less&nbsp;stability,&nbsp;but higher pay; another may value stability at the cost of slightly lower compensation. Hard to generalize, but knowledge of multiple technologies and areas certainly helps.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">How do you view salaries in this role compared to the IT market?&nbsp;<\/h3>\n\n<p>I know technologies with much higher pay, and some with&nbsp;lower pay. Due to the current hype, AI salaries have&nbsp;largely leveled&nbsp;with the average in more mature technologies. I believe salaries in this role are decent, but a good specialist will be properly compensated regardless of technology.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">What is a bigger motivator: salary or project impact?&nbsp;<\/h3>\n\n<p>Everyone would lie if they claimed salary&nbsp;doesn\u2019t&nbsp;affect motivation. But at some point, money stops being the main driver. Much more important are agency, recognition, real influence on processes, and development in areas of personal interest. For me,&nbsp;it\u2019s&nbsp;important that I have a real impact and am appreciated for it.&nbsp;<\/p>\n\n<h2 class=\"wp-block-heading\">Future of&nbsp;Machine&nbsp;Learning&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">What technological trends will&nbsp;impact&nbsp;the work of an AI\/ML Architect?&nbsp;<\/h3>\n\n<p>In the coming years, the trend will be developing AI solutions that truly solve problems and generate real value, rather than being empty buzzwords that someone plays with for a few days and forgets. More mature tools, frameworks, and models will&nbsp;emerge. The market has accelerated significantly, and planning horizons have shortened from years to quarters. I believe the architect\u2019s role will become increasingly necessary and crucial &#8211; not only as an \u201cAI Architect\u201d in a narrow sense, but as someone who sees the bigger picture and understands more than one element of the entire ecosystem. A strongly underestimated trend is AI security. We are already seeing new attack vectors targeting GenAI-based systems, and AI security is becoming a new specialization within cybersecurity.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">Is GenAI changing how architecture is designed?&nbsp;<\/h3>\n\n<p>Yes, and this can be broken down into several factors. On one hand, GenAI and large language models have significantly accelerated work&nbsp;&#8211;&nbsp;making it easier to explore solutions, search documentation, and discuss approaches. Often, it helps me deepen topics I&nbsp;wasn\u2019t&nbsp;familiar with before. But it also brings risks&nbsp;&#8211;&nbsp;model hallucinations, incorrect information, or confirmation of bias. If we approach LLMs poorly, we may get answers that agree with even the worst idea. I see this as a real threat. Sometimes roles can reverse,&nbsp;rather than verifying, thinking, and discussing; we become people who accept generated solutions.&nbsp;It\u2019s&nbsp;like someone&nbsp;wrote&nbsp;a paper for you; what good is a perfect grade if you cannot answer questions about it later? I use different models and applications for different problems, which helps speed up work and focus on important tasks rather than repetitive ones. The key is not to switch off critical and creative thinking. These tools should help us, not think for us.&nbsp;<\/p>\n\n<h3 class=\"wp-block-heading\">What&nbsp;regulatory and&nbsp;ethical&nbsp;challenges&nbsp;do AI&nbsp;projects&nbsp;face?&nbsp;<\/h3>\n\n<p>This topic grows every year. We must consider the AI Act, GDPR, ethics regarding model usage, and how\/where we store and process data. Practical questions are numerous\u2014from where to store training data, to anonymizing it, to documenting decisions made by models. Regulations organize the market and force us to think about things we should already consider.<\/p>\n\n<h2 class=\"wp-block-heading\">AI\/ML Architect&nbsp;work&nbsp;at&nbsp;Sii&nbsp;Warsaw&nbsp;<\/h2>\n\n<h3 class=\"wp-block-heading\">What&nbsp;makes&nbsp;working&nbsp;at&nbsp;Sii&nbsp;different&nbsp;from&nbsp;other&nbsp;companies?&nbsp;<\/h3>\n\n<p>I can only speak for myself, as it may vary by team. But one thing is common: if you have passion and motivation, Sii provides an environment for growth. I have worked here for 9 years. I was lucky to have great managers who saw my potential and invested time in my development.<\/p>\n\n<h3 class=\"wp-block-heading\">What&nbsp;kinds&nbsp;of AI&nbsp;projects&nbsp;do&nbsp;you&nbsp;undertake?&nbsp;<\/h3>\n\n<p>Sii consists of various Competence Centers, including the AI Competence Center, collaborating with many clients, so the company-wide portfolio is broad. Our team executes IT projects using AI\u2014from NLP, Computer Vision, and recommendation systems to GenAI and machine learning, including agent-based systems. We have several large AI projects and a dozen smaller ones.<\/p>\n\n<h3 class=\"wp-block-heading\">How&nbsp;mature&nbsp;are&nbsp;AI&nbsp;projects&nbsp;at&nbsp;Sii?&nbsp;<\/h3>\n\n<p>Sii has been implementing AI projects for many years. Many are deployed in production, and we are now seeing rapid growth in GenAI-based solutions. Some projects remain in PoC when technology isn\u2019t mature enough or data doesn\u2019t answer the questions. But new opportunities continually arise. We are constantly developing, and the AI\/ML field is changing dynamically.<br><\/p>\n\n<h3 class=\"wp-block-heading\">What&nbsp;do&nbsp;you&nbsp;value&nbsp;most&nbsp;about&nbsp;working&nbsp;at&nbsp;Sii?&nbsp;<\/h3>\n\n<p>The people and the atmosphere. After so many years, I know this is not empty words. I started as a C++ developer and now I\u2019m an AI Architect. This path was possible because someone gave me a chance and space to learn. Even though the company has grown, the atmosphere and approach to people haven\u2019t changed.<\/p>\n\n<h3 class=\"wp-block-heading\">In one&nbsp;sentence,&nbsp;why&nbsp;is&nbsp;it&nbsp;worth&nbsp;being&nbsp;an&nbsp;AI\/ML Architect&nbsp;at&nbsp;Sii?&nbsp;<\/h3>\n\n<p>It\u2019s a challenging role full of opportunities, bringing great satisfaction if you are motivated and truly passionate about what you do.<\/p>\n\n<p><\/p><\/div>","protected":false},"excerpt":{"rendered":"","protected":false},"author":131,"featured_media":138637,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"footnotes":""},"tags":[5742,6193,6201],"class_list":["post-138649","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","tag-artificial-intelligence","tag-career-paths","tag-interview"],"acf":[],"aioseo_notices":[],"featured_media_url":"https:\/\/sii.pl\/wp-content\/uploads\/2026\/03\/1920x740_cover-www_inside_the_role_ai_ml.jpg","category_names":[],"_links":{"self":[{"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/posts\/138649"}],"collection":[{"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/users\/131"}],"replies":[{"embeddable":true,"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/comments?post=138649"}],"version-history":[{"count":3,"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/posts\/138649\/revisions"}],"predecessor-version":[{"id":140376,"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/posts\/138649\/revisions\/140376"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/media\/138637"}],"wp:attachment":[{"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/media?parent=138649"}],"wp:term":[{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sii.pl\/en\/wp-json\/wp\/v2\/tags?post=138649"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}