We are so excited about this case study, as we strongly believe in Clinical Trials Transformation Initiative (CTTI)’s mission to develop and drive adoption of practices that will increase the quality and efficiency of clinical trials.

Think clinicaltrials.gov, but with all structured information in a database. This is AACT and it is an immensely valuable resource for intelligence in clinical trials. Before AI entered this space, you still needed to be a data scientist to access this resource.

Here's a peek at what it can do - an analysis of bispecific antibody patents: claude.site/artifacts/d9...

Want to try it yourself? Check out our GitHub repo for source code and setup instructions: github.com/navisbio/mcp...

⚡️ Quick note: While public BigQuery databases may be a bit dated, we're developing proprietary databases and tools for AI-powered translational research. Need access to current patent data, clinical trials, and more? Let's connect!

Thank you! Yes, exactly. It’s really fun to explore the database this way.

The numbers comprise industry-sponsor-led interventional trials assessing drugs, biologics &
genetic therapies and were retrieved from AACT/clinicaltrials.gov using our AI agent workflow.

𝐄𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐜𝐲 𝐯𝐬. 𝐦𝐚𝐧𝐚𝐠𝐢𝐧𝐠 𝐜𝐨𝐦𝐩𝐥𝐞𝐱𝐢𝐭𝐲 𝐯𝐬. 𝐝𝐞 𝐧𝐨𝐯𝐨 𝐢𝐧𝐬𝐢𝐠𝐡𝐭𝐬

However, value propositions beyond efficiency were sought - managing complexity, improving decisions, generating novel insights. Challenge: benefits take years to manifest and are hard to disentangle from other variables.

𝐅𝐢𝐧𝐝𝐢𝐧𝐠 𝐭𝐡𝐞 𝐫𝐢𝐠𝐡𝐭 𝐀𝐈 𝐯𝐚𝐥𝐮𝐞 𝐩𝐫𝐨𝐩𝐨𝐬𝐢𝐭𝐢𝐨𝐧 𝐢𝐧 𝐛𝐢𝐨𝐩𝐡𝐚𝐫𝐦𝐚 𝐢𝐬 𝐜𝐡𝐚𝐥𝐥𝐞𝐧𝐠𝐢𝐧𝐠

Opinions varied on efficiency and cost savings, with skepticism about pure efficiency gains. Many said incremental gains were too small vs technical and cultural change required.

Of note, maximum message length can become a limitation when assessing several targets at once and integrating many metrics in the analysis. For more complex workflows we envision separate agent workflows orchestrated from one main conversation.