A 10-minute research presentation followed by a long Q&A (approx. 45-60 mins total). The panel consisted of three scientists with heavy Statistics and Chemical Engineering backgrounds. The vibe is more "Applied Data Science" than "Computational Physics." Pros: - Respectful environment; they give you ample time to finish your presentation before asking questions. - Interesting look into how traditional giants are adopting RAG and LLMs for internal documentation. Cons: - Heavy emphasis on specific domain fit (e.g., Polymers). If your PhD is in Inorganic/Crystals, the gap might be a dealbreaker regardless of ML skill.

I was invited to the location where together with a group of other candidates we had a comp y presentation followed by interviews and solving a business case. Prior to that I sent my resume. The interview was nice, people interviewing me were informed and asked good questions.