AI Based Patent Search For Chemical Compound & Drug Discovery
summary
Builds an AI pipeline that codifies chemical formulas into embeddings and scores probability of patent infringement against databases from USPTO Google Chem Intelligence Potion, using pattern‑matching and prior litigation signals to output a risk score instead of a binary yes/no. Targets chemists and patent attorneys, ingests multimodal data, flags novel versus similar compounds, and produces explainable traces for legal review. Prototype exists after two months of work and is slated for internal use and later commercialization.
AI-written summary · about
chapters (7)
- 0:00Introduction to AI in drug discovery
- 1:55Problem of patent infringement with AI
- 3:42Challenges in AI-based patent search
- 4:44Data collection and AI training process
- 6:26AI as augmentation for chemists and attorneys
- 7:30Government role in ethical AI
- 8:35Q&A on Aiden’s technical journey
related talks
41mSimilarity Search in Patents via modern Artificial Intelligence
16mBeyond Patterns: The Case for Physics-Grounded AI in Drug Discovery
30mAI in Drug Discovery
35mCan AI solve the biggest problem in drug discovery? Time!
22mDeconstructing the US Patent Database
47mAI for the Proper Use of Medications
08mEngineering the Future of Pharma: AI-Driven Drug Discovery
36mFormatting Biological Big Data to Enable (Personalized) Systems Pharmacology