Case Study: A Discovery Group Integrates Alchemy Probability Data into Their RAG and Discovery Workflows

A life sciences discovery group was building AI-assisted pipelines for drug and materials discovery. They had in-house RAG and query systems; they needed a comprehensive probability database—millions of element and compound combinations, with methods and probabilities—that was documented and assembled for RAG and other AI workflows. They did not want to build the underlying methodology from scratch; they wanted a complete, documented asset they could query, rank, and prioritize with.

The situation. The group had evaluated several data products. Many were narrow or not structured for RAG. They wanted something that underpinned cure discoveries and chemistry (they had seen the Chemical Cooker and cure discovery products from the same portfolio) and that could also be used as a standalone data layer. The package had to include the database, documentation, and assembly suitable for their existing tools.

What they did. They licensed the Alchemy Probability Data product from Christopher Gabriel Brown: comprehensive probability database, millions of combinations, methods and probabilities, documentation, and RAG-ready assembly. They integrated the database into their discovery pipelines and ran a series of queries to validate coverage and relevance for their targets. They did not use the Chemical Cooker or cure products in this engagement; they used Alchemy Data as a standalone handoff. They confirmed that base delivery was one finished product copy and that IP was not transferred unless separately agreed.

Outcome. The group put the database into production for two internal programs—one drug discovery, one materials—and reported faster candidate ranking and better alignment with their existing methodology. They noted that the same data and methodology appeared in other products in the portfolio, so the full scope of applications was broader than their immediate use. No external publication or product was launched in the case study period; the outcome was an operational data asset and a clearer view of how the portfolio fit together.

Takeaway. Probability data is valuable as a standalone handoff when it is complete, documented, and assembled for the way you work (e.g. RAG, query, ranking). The Alchemy Data product can power discovery across chemistry, materials, and life sciences without requiring you to take the Chemical Cooker or cure products. The case study showed that the data layer alone was sufficient for one buyer’s needs.

View the full portfolio of 18 technology products →
Companion piece: Product Insight — Alchemy Probability Data

Masonry view