Key results
The company
Tangram Therapeutics
tangramtx.comBiotech platform for computational drug discovery and RNAi medicines.
Result highlights
- Up to 50x faster target assessment
- 300x increase in data processing volume
The story
A UK-based biotech company merging biology and computation to accelerate the discovery of RNA interference (RNAi) medicines for diseases with significant unmet needs.
Traditional drug discovery timelines were prohibitively slow, with document processing taking weeks and target-indication assessments requiring a full quarter. Scientists struggled to synthesize massive amounts of proprietary and public data to identify novel gene targets using legacy deep learning models.
The company developed an agentic AI platform on AWS that orchestrates multiple Large Language Models via Amazon Bedrock to analyze over 1,000 biological datasets. The system uses Retrieval Augmented Generation and document processing pipelines to synthesize internal lab results and public data, while embedding subject matter expertise directly into prompt design. A modular architecture allows the team to continuously benchmark and swap in new models as technology evolves.
Quotes
“We need the best of the best LLMs at any given time. Everything we’ve done is modular, extensible, and changeable, so that as we get increasingly better technology, we can just plug it in.”