Scientist LLM's and agentic AI

About this position

Conflicts, crime, and subversive activities threaten our security worldwide. To counter these threats, TNO conducts innovative research and works on applying these insights with our security partners. In doing so, we support organizations such as the Ministry of Defence and the Police in making the best-informed decisions and carrying out (semi-)automated actions they can rely on.

We do this by conducting applied research, generating innovative ideas, and translating them into practical solutions. For example, in the use of intelligence for Defence, the development of decision support models for the Public Prosecution Service, and the application of NLP to analyze criminal communications. In this role, you will join the Intelligence & Decision Support research group. This driven team of around 40 colleagues brings together a wide range of expertise and backgrounds and is part of the Defence, Safety & Security unit.

What will be your role?

As a scientist, you contribute to research projects focused on LLMs and agentic AI for decision support within the defence and security domain. Working in multidisciplinary teams, you engage with various research questions, helping to shape the approach and methodology needed to answer them. Based on scientific research, existing methodologies, and best practices, you and your team develop concepts, with modelling and programming around LLMs forming an essential part of your work. The result may, for example, be a proof of concept, where you consistently keep in mind the impact of the tools you develop on the daily operations of our partners. You communicate project results to our partners and act as a valuable sparring partner for colleagues.

Examples of projects include developing an LLM concept for decision support within the Police, creating a proof of concept for mission planning in Defence, contributing to a Dutch LLM, and making the feasibility of agentic AI tangible for security partners.

At TNO, there is ample opportunity to deepen your expertise in new AI tools, attend conferences, and explore new research directions. There are also opportunities to initiate internal research lines and contribute to project acquisition and proposals for collaborations. In addition to project work, you can take on various side activities aligned with your interests and development goals, such as coordinating internships within the department, organizing content-driven or social events, and mentoring new colleagues.

What we expect from you

You are driven to contribute to the military and security domain through your work and are eager to learn more about it, so your work increasingly aligns with the needs of our partners. You handle sensitive information with care and are interested in, for example, participating in exercises or working on-site with a partner for a period of time to build valuable domain knowledge. With your curiosity, you actively seek out the information you need for your assignments. You are analytically strong and able to help translate scientific insights into practical applications. You are comfortable working on multiple projects and within different teams simultaneously and can switch effectively between them. You enjoy collaboration, leveraging your communication skills, while also being able to plan your work independently and confidently put forward proposals and take the lead. As a result, you are well equipped to present conclusions in written form, presentations, diagrams, and data visualizations, as well as facilitate expert workshops.

You are enthusiastic about working with AI, particularly LLMs, and are eager to further develop your expertise in this area in the coming years. In addition to prompt engineering, you are interested in topics such as fine-tuning, explainability, robustness, adversarial attacks, and other aspects that con