I’m Bastien Hottelet, a Master’s student in Machine Learning & Data Science at Université Paris Cité (graduating September 2026) and an applied AI researcher passionate about building robust, truth-seeking AI systems. My work bridges the gap between academic research and real-world deployment—focusing on how large language models reason, verify, and adapt to imperfect information. I currently work at Safran Aircraft Engines as a Data Scientist in Generative AI, where I develop agentic systems and scalable NLP pipelines. Alongside, I’m working on two independent research projects: IMMUNE-Bench, a framework for training and evaluating agents on epistemic robustness, and an open-source AI knowledge graph that helps researchers structure and connect papers automatically.

Agentic AI and reinforcement learning, truthfulness evaluation, bias and alignment studies, and AI tools that improve scientific research and knowledge management.

IMMUNE-Bench: framework for training and evaluating LLM agents on factual robustness. Research Graph Agent: Dockerized app linking papers via AI-inferred relations.

I am a Java developer and recent Computer Science graduate (BUT) who completed an apprenticeship with CGI France. Passionate about full-stack development, AI, and data science, I’m now self-learning cybersecurity. I have worked on innovative projects like a GPT-2 medical chatbot fine-tuned with LoRA and automated VM configuration systems, I combine technical expertise with a drive for continuous learning. My goal is to master cybersecurity while contributing to cutting-edge tech solutions.

Artificial intelligence, data science, cybersecurity, full-stack development

Currently developing a self-learning cybersecurity project.