AI Advances: Stay Up-to-Date

2025. What a year. Honestly, I’m wrapping up this year with a feeling that’s hard to put into one word… but if I had to choose: gratitude.

Hi there, 2025 was the year AI became unavoidable. 2026 will be the year it becomes uncontrollable, in a good way, mostly. Below are my predictions for where AI is heading next year. Not the safe ones. The real ones. Direct, sharp, and maybe a bit sarcastic. Just how we like it.

If I'm going to share the story of our legal tech venture, I need to speak from the heart. So here it is a candid, personal take on why we are not chasing the usual hype. We’re in an early-stage stealth mode legal tech startup, and our guiding principle is simple: speed isn’t the goal in legal, but precision, security, and data quality are. In a field built on trust and accuracy, moving fast and breaking things just isn’t an option.

When I first heard about EU-INC, I felt two things at the same time. I was optimistic because the idea of a single European company form could finally let founders incorporate once and scale across the entire continent. I was also realistic, because this is Europe with its twenty seven languages, twenty seven tax codes and twenty seven opinions.

Everywhere I look in tech, newsletters, conferences, coffee chats, I’m bombarded by new “Ops” talk. DevOps, MLOps, LLMOps, AIOps… It sometimes feels like there’s a new flavor every quarter. And if you’re like me, you might wonder if we’re reaching real progress, or just rebranding the same challenges under new names.

I often come here to discuss AI, models, protocols, and the latest breakthroughs. But let’s be honest: none of that matters if you haven’t actually started applying AI in your work or company. The real value comes from moving beyond theory and getting your hands dirty. So today, I’m sharing my practical guide to help you start using AI the right way, making sure you don’t miss any crucial steps on the path to future success.

As both a father and an AI leader, I’m constantly reflecting on how quickly artificial intelligence is transforming the way our children learn. My daughter is already interacting with AI tools, sometimes more intuitively than I do. It’s both exciting and daunting.

In the rapidly evolving AI landscape, the rise of autonomous agents demands robust, standardized protocols to enable seamless collaboration, data access, and orchestration. If you’ve been following the recent LuminaTalks podcast episode on ACP vs A2A vs MCP, you know these protocols are shaping the future of multi-agent AI systems.

When done right, AI can be an incredible catalyst for speed, efficiency, and innovation.

When people think about failed AI projects, they often assume the issue must be technical. They blame the model, the algorithm, the complexity of the problem. But in my experience, that’s rarely where the real failure starts.

Imagine you're watching a symphony, dozens of instruments playing in perfect coordination. The music is flawless, until one violin plays a wrong note. Suddenly, the illusion breaks. That one misstep doesn’t just disrupt the sound, it disrupts trust in the entire performance. That’s what building AI systems feels like today. That’s what a lot of AI implementation looks like today.

I used to think compliance was something you dealt with later. After product. After sales. After fundraising. But the more our business grows, and the more AI tools we put into play, the more I’ve realized: you can’t afford to think like that.

Giving AI agents raw data without context is like handing someone a map with no labels. Sure, they can see the roads, but they have no idea where they’re going.

Picture this: You’re planning a long, ambitious road trip. Excitement is high, the destination is clear, and you can’t wait to hit the road. But before you start the engine, there are a few things you must do: check the engine, ensure the brakes work, map your route, and pack an emergency kit. These aren’t just optional steps—they’re essential to arriving safely.

Data governance reminds me of plumbing. When it works, no one thinks about it. But when it fails, everything else stops working—and the clean-up is painful and expensive.

We rarely talk about it in meetings or strategy decks. But everyone building something meaningful knows it’s there: Integration is painful! It slows us down. It eats dev hours. It bloats budgets. And most of all—it chips away at momentum.

Let’s face it—most companies are buried under a mountain of disconnected software. You start with one tool to fix a small problem. Then another. And another. Suddenly, you’ve got ten platforms—each great on their own—but they barely speak to each other. And now your team spends more time connecting tools than actually using them.

Twelve months ago we onboarded a fast‑growing retail marketplace that had just crossed 100 micro‑services. Reporting dashboards were breaking daily, analysts were emailing CSVs back and forth, and every new feature release meant another sleepless integration weekend. When we refactored their data stack around a modular “Lego” ecosystem their nightly load time dropped from 9 hours to 45 minutes and their team shipped a new buy‑now‑pay‑later feature in two sprints instead of six. I’ve seen the same pattern repeat across fintech, health‑tech, and SaaS: modular data wins every time.

Throughout my journey as CEO of Summ.link link, I've encountered countless business leaders and entrepreneurs who frequently hear terms like "data governance" but rarely grasp its true importance until it becomes a significant issue. Today, I want to share why robust data governance isn't merely about compliance—it’s a strategic foundation for sustainable and scalable growth.

AI is everywhere now—chatbots, data analytics, automation tools—but here’s the problem: most of these systems don’t actually work together. It’s like having a team where no one speaks the same language.

AI is no longer a future consideration—it is actively reshaping how companies train, upskill, and retain employees. As automation transforms industries, workplace learning is shifting from static programs to AI-driven, data-backed strategies that improve efficiency and outcomes.

AI and data regulations are changing at lightning speed, and the stakes are high for businesses and consumers alike. In this edition, I’ll break down the latest updates on the AI Act and GDPR—what’s changing, why it matters, and how it will shape the future of AI.

Artificial intelligence is no longer an emerging trend—it’s the driving force behind business innovation, efficiency, and competition. With AI adoption accelerating across industries, the demand for AI professionals is shifting from general knowledge to specialized expertise.

Every business leader I talk to faces the same challenge: there aren’t enough skilled employees, and the ones they have are stuck handling repetitive tasks. Customer support teams answer the same questions day in, day out. Doctors spend hours writing reports instead of seeing patients. Legal teams manually review contracts for minor updates. These inefficiencies slow businesses down and cost millions in lost productivity.

AI is evolving fast, and two game-changing concepts are making waves: Vertical AI and Multi-Agent Systems (MAS). These aren’t just cool tech upgrades—they’re reshaping how industries work, streamline processes, and make smarter decisions.

As AI continues to take center stage in business transformation, one thing has become painfully clear: too many companies jump to solutions before fully understanding their actual problems. And that’s where most AI initiatives go wrong.

I recently came across an exciting pre-print on ArXiv where researchers introduced a new technique called Cache-Augmented Generation (CAG). This approach has the potential to reduce the reliance on RAG (Retrieval-Augmented Generation) and eliminate many of its drawbacks. Instead of retrieving information from an external knowledge store, CAG pre-loads all relevant knowledge directly into the extended context of a large language model (LLM) and uses it to answer queries during inference. What’s even more interesting is that when applied to long-context LLMs, the results suggest that CAG not only performs on par with RAG but in some cases even outperforms it across various benchmarks. Let me take you through this fascinating technique, how it works, and how it compares to RAG.

Om eerlijk te zijn, ik durf nog niet te zeggen of het een evolutie of revolutie wordt. Maar dat het een stap vooruit zal zijn, staat vast. Hoe groot die stap zal zijn, zal de komende jaren blijken. Dat Vertical AI of Multi-Agentic Agents de volgende grote innovatiegolf vormen na SaaS, lijkt zeker. Hoewel we ons nog in de beginfase bevinden, is het duidelijk dat deze technologie niet alleen de manier verandert waarop bedrijven werken, maar ook een diepgaande impact heeft op efficiëntie, kostenbesparing en het versnellen van bedrijfsprocessen. En dat is nog zonder het enorme potentieel voor talentwerving te noemen, vooral als men niet internationaal wil rekruteren.

In the evolving landscape of artificial intelligence, Azure OpenAI’s GPT-4o Realtime API is transforming how businesses communicate. By combining VoiceRAG (Voice-based Retrieval-Augmented Generation) with AgenticRAG and Knowledge Graphs, businesses can create intelligent, context-aware voicebots that not only interact in real-time but also make informed decisions based on structured knowledge.

Laten we eerst beginnen met dit: zelfstandig worden is de beste keuze die ik ooit heb gemaakt 🚀. Na 6,5 jaar voor een werkgever te hebben gewerkt, was ik op zoek naar een nieuwe uitdaging en wilde ik proeven van verschillende aspecten van het ondernemerschap. Maar wat ik ontdekte, is dat het een echt mijnenveld 💥 kan zijn!

Het voorstel van #OpenVLD om 10% van de ambtenarenjobs te laten verdwijnen door AI roept verschillende vragen en bedenkingen op. Laten we eens kijken naar de cijfers die je aanhaalt. Volgens recente statistieken werkt 18.4% van de werkende mensen in België bij de overheid, wat neerkomt op ongeveer 775.490 mensen. Als we dit aantal met 10% willen verminderen, zouden we uitkomen op 697.941 ofwel 77.549 FTE's minder dan vandaag! Houden we rekening met een gemiddeld overheidsloon in België van iets boven de €4.000 (Statbel), zou dit een besparing opleveren van €3.7 miljard per jaar.

Hold onto your seats because we've got some thrilling news in this edition of our AI Newsletter! The European Union has just thrown open the doors of its supercomputers (1) , and it's about to make some major waves in the world of AI. Plus, we'll spill the beans on how this fantastic opportunity is set to supercharge HR-AI companies like Summ.link and Centr.jobs.

In today's rapidly evolving world, the impact of AI on jobs is a topic that's on everyone's mind. You often hear or read "AI is taking jobs!", but I see it different, more positive: AI is reshaping careers!

As a father and entrepreneur, I strongly believe that introducing AI education to every child is an essential step in preparing them for the future. Literacy and mathematics have long stood as the core pillars. But the world does not stand still, and we believe that it is past time for a new core pillar: computer science and AI coding.

In today's data-driven world, where freedom, innovation, and people are paramount, organizations must prioritize information security and establish trustworthiness. Recently, our team had the privilege of interviewing DQS, a renowned ISO certification partner, to delve into the significance of achieving ISO 27001 certification and discover strategies for leveraging this achievement to excel in the market. Through this article, we aim to share the valuable insights gained from the conversation, highlighting Summ.link's remarkable accomplishment and emphasizing the immense value that ISO 27001 certification brings to the organization's growth and reputation, while staying true to its core values of freedom, innovation, and people.

In today's rapidly changing business landscape, companies are constantly seeking ways to optimize their operations and make data-driven decisions. HR and People Analytics have become crucial tools for organizations looking to attract and retain top talent, increase productivity, and improve overall employee satisfaction. With the emergence of Data Science as a Service (DSaaS), companies now have the opportunity to leverage cutting-edge technologies and data analytics to gain deeper insights into their workforce. In this article, we will explore the benefits of investing in DSaaS for HR and People Analytics, and how it can transform the way you manage your human capital.

In 2017, the Economist published a report titled “The world’s most valuable resource is no longer oil, but data.” Following this, many industry experts and publications hailed “data” and “data-driven decision making” as the next big thing in business. As more and more companies and institutions began to realize the power of using data, there emerged the phenomenon of a “data-driven organization.

Artificial Intelligence (AI) has become an integral part of our daily lives and is used in various industries such as education, healthcare, automobile, manufacturing, and law enforcement.

In this newsletter, we will be exploring the topic of explainable artificial intelligence (XAI). XAI refers to artificial intelligence systems that are able to provide explanations, insights, and reasoning behind their decisions and actions. These systems are designed to be more transparent and understandable to humans, allowing users to understand how the system arrived at its conclusions and to identify any potential biases or errors.

Welcome to our sixth newsletter on people analytics! In this edition, we will focus on the impact of people analytics for HR and talent acquisition teams. People analytics can be a powerful tool for HR and talent acquisition professionals, providing insights into the skills, performance, and potential of employees and candidates. By using data and analytics to inform decision-making, HR and talent acquisition teams can improve the efficiency and effectiveness of their efforts, leading to better outcomes for the organization and its employees. In this newsletter, we will explore the ways in which people analytics can benefit HR and talent acquisition teams, including increased efficiency, improved accuracy and consistency, enhanced personalization, and greater adaptability. We will also discuss the ethical and legal considerations that need to be taken into account when using people analytics in these contexts.

Welcome back to our newsletter on people analytics! In this edition, we will focus on the potential applications of people analytics, as well as the ethical and legal concerns that need to be considered. People analytics can be used to improve a wide range of HR and talent management practices, including talent acquisition, performance management, employee engagement, and succession planning. However, the use of data in people analytics also raises a number of ethical and legal concerns, including issues related to privacy, bias, and discrimination. It is important for organizations to carefully consider these issues and put appropriate safeguards in place to ensure responsible and ethical use of data in people analytics efforts.

Welcome back to our newsletter on people analytics! In this edition, we will focus on the potential applications of people analytics, as well as the ethical and legal concerns that need to be considered. People analytics can be used to improve a wide range of HR and talent management practices, including talent acquisition, performance management, employee engagement, and succession planning. However, the use of data in people analytics also raises a number of ethical and legal concerns, including issues related to privacy, bias, and discrimination. It is important for organizations to carefully consider these issues and put appropriate safeguards in place to ensure responsible and ethical use of data in people analytics efforts.

Welcome back to our newsletter on people analytics! In this edition, we will focus on the potential applications of people analytics, as well as the ethical and legal concerns that need to be considered. People analytics can be used to improve a wide range of HR and talent management practices, including talent acquisition, performance management, employee engagement, and succession planning. However, the use of data in people analytics also raises a number of ethical and legal concerns, including issues related to privacy, bias, and discrimination. It is important for organizations to carefully consider these issues and put appropriate safeguards in place to ensure responsible and ethical use of data in people analytics efforts.

Welcome to the first newsletter on people analytics! In this edition, we will focus on what people analytics is and the most important data sources to solve people analytics issues. People analytics is the practice of using data and analytics to understand and optimize the performance and potential of individuals and teams within an organization. It involves collecting and analyzing data about people-related factors such as skills, performance, engagement, retention, and more, and using this data to inform decision-making and drive improvements in HR and talent management practices. Some common types of data sources within people analytics include HRIS data, surveys and assessments, social media data, external labor market data, performance data, exit interview data, and behavioral data.

Artificial intelligence has come a long way in recent years, and one of the most exciting areas of development is the field of gen-AI. Gen-AI, or AI that is generated or synthesized rather than explicitly programmed, has the potential to revolutionize the way we think about artificial intelligence and how it is used. In this text, we will explore the concept of gen-AI, examining its impact on society and the economy, as well as how it is trained and created. We will also delve into the world of language models, and consider why they exist and the potential revenue models associated with gen-AI. With so much to learn about gen-AI, let's dive in and discover the many ways it is shaping the future of artificial intelligence.