Enterprise Superintelligence

Executive Summary Agentic workflows are AI-driven systems designed to reason, plan, and act autonomously across complex tasks and environments. Unlike prompt-based interactions or deterministic automation, they coordinate multiple AI agents, tools, and feedback loops to achieve high-level goals under uncertainty. This article introduces the core components of an agentic workflow, explains how autonomous AI agents operate within them, and clarifies when agentic approaches are appropriate for production AI systems.

Executive Summary Many essential consumer goods, from life-saving chemotherapy drugs to the fragrances and flavors in everyday products, come from compounds found in plants. These are cornerstones of multi-billion dollar markets where consumer demand for sustainable, natural products is growing at an unprecedented rate. However, the methods used to acquire them are often anything but sustainable. The development of more sustainable approaches, such as Microbial Cell Factories, presents complex engineering challenges, leading to costly and slow R D. Active Learning can revolutionize the traditional research cycle by delivering a faster, cheaper and more intelligent R D process, paving the way for a sustainable future.

Executive Summary The field of Artificial Intelligence is in a constant state of evolution. For years, the focus was on building a single, powerful model capable of tackling any task thrown its way. But as the complexity of our problems grows, it's becoming clear that a lone genius, no matter how brilliant, can be outmatched by a well-coordinated team of specialists. This is the paradigm shift that brings us to multi-agent AI systems. A Multi-Agent System (MAS) is a team of specialized AI agents that collaborate to tackle complex problems. Use a multi-agent system when your challenge involves multiple tasks, dynamic environments, or specialized expertise that require agent coordination and agent communication. For simpler or well-defined workflows, a single intelligent agent or structured agentic workflow is often a better, faster, and more cost-efficient choice. The golden rule is to use the simplest setup that effectively solves your problem.

AI is redefining software development — faster than many expected. In the article by De Morgen, our Senior Engineer Niels Rogge shares his perspective on the rapid rise of AI-powered software development and the explosion of so-called “vibe coding” — building applications largely generated by AI. Thanks to advances in AI coding agents, tasks that once took weeks can now be completed in days. But speed isn’t everything. As Niels puts it:

Ghent-based ML6 to make Belgium’s Federal Public Service of Justice more transparent and efficient.

AI in protein folding: a virtual playground Protein folding — the way amino acid chains fold into 3D structures — has long been one of the hardest problems in biosciences. With breakthroughs like DeepMind’s AlphaFold, recently awarded the Nobel Prize for Chemistry, predicting these structures is now faster and more accurate than ever.