Our software engineer Dumindu de Silva completed his master’s thesis

The work, carried out in collaboration between the University of Turku (UTU) and Fatman, addresses a very practical industrial challenge: Machinery troubleshooting often relies on extensive technical documentation that can be slow to navigate in time-critical situations.

To address this, the thesis designed and implemented a Retrieval-Augmented Generation (RAG) based chatbot that combines document retrieval with generative AI. This allows users to ask questions in natural language and receive answers grounded in technical manuals. The solution was also implemented and tested as a working product.

The work supports Fatman’s ongoing development of AI-powered tools aimed at making industrial knowledge more accessible and usable in practice, while also tackling real-world engineering challenges such as scalable architecture, multi-tenant separation, and system integration.

We also warmly acknowledge the collaboration with the University of Turku (UTU), which supported the academic side of the work. At Fatman, we highly value research that connects academic insight with practical industrial impact and this thesis is a strong example of that.

Big congratulations to Dumindu on this great achievement!

And for all interested in the details, here’s the thesis link:
https://www.utupub.fi/items/75e65b4e-e073-4167-896d-49f066f60b12