Tim de Boer

Delphyr

AI Engineer

Building Medical AI at Delphyr, a clinical assistant that helps healthcare professionals reclaim their time. Owning the guardrails & evaluation stack: prompt design, retrieval, hallucination detection, citation verification, and end-to-end scenario testing for real clinical use.

Future Facts Conclusion

AI Engineer

Built RAG-based chatbots for enterprise knowledge bases, achieving up to 75% time savings for customer service teams. Co-developed the first Conclusion voicebot, focusing on prompt engineering, sentiment-aware tone, and compliance with the AI Act and GDPR.

• Production RAG chatbot for a national customer-service operation
• Voicebot prototype with cloned voice and live sentiment adaptation
• Atradius IFRS17 case: migrated a legacy SAS data pipeline to Databricks & PySpark, cutting runtime from 24+ hours to under 2

VU Amsterdam, MSc Thesis

Brain-Computer Interface researcher

Six-month master's project building an end-to-end real-time motor-imagery BCI on a consumer EEG headset (g.tec Unicorn). Pipeline included artefact rejection, CSP/Riemannian feature extraction, deep transfer learning with EEGNET, and a live integration that let subjects play Space Invaders with their thoughts.

MSc Artificial Intelligence

Vrije Universiteit Amsterdam · cum laude

Graduated cum laude (8.6/10). Master research project carried out in Elche, Spain, in the lab of Professor Jose M. Azorin: "Deep transfer learning for real-time upper-limb motor imagery decoding with a low-cost EEG device." Graded 9.0/10.

BSc Human Movement Sciences

Vrije Universiteit Amsterdam

Biomechanics, motor control and exercise physiology, the foundation behind my interest in artificial intelligence and the human body. Bachelor project: "Assessing the validity of 2D-ultrasound linear extrapolation techniques for measuring biceps femoris long-head muscle architecture." Graded 9.0/10.