Evaluating Agentic AI Frameworks for Real-World Drug Design Tasks
By Anton Sinitskiy, with students
🔗 linkedin.com/in/antonsinitskiy
Research Focus: Applied AI, Drug Design / Pharma
Type: Other Student Research
What research challenge does this work address?
Sinitskiy and his students ran evaluations of how well advanced agentic AI frameworks perform on real-world tasks, with a focus on drug design.
What inspired this project?
The team wanted its results to have practical value for industry and for CPS alumni entering AI-driven fields.

What is the outcome and impact?
The team concluded that using AI for tasks like drug design is still more of an art than a well-defined technique, and provided specific examples of what works and what doesn’t. The findings should be of high interest to industry practitioners working to apply AI in this space, and should help prepare CPS alumni to be even more ready for their future work in the field.
Findings are documented across a series of bioRxiv preprints.
Read preprint 1 on bioRxiv.
Read preprint 2 on bioRxiv.
Read preprint 3 on bioRxiv.