Inductive Bio Unveils Indy, an AI Agent for Medicinal Chemists

With an 89% accuracy rate in head-to-head benchmarks, the newly launched Indy AI assistant is setting a new performance bar for drug discovery. By automating routine data analysis and decision-making tasks, the tool aims to double the productivity of medicinal chemists currently bogged down by operational bottlenecks.

Corp and Tech·September 15, 2026·reads: 3,417

Inductive Bio developed Indy by training the model on the workflows of expert drug hunters to handle the complex, time-consuming requirements of medicinal chemistry. Unlike general-purpose models, Indy integrates directly with a program’s internal data and computational tools, including Inductive’s proprietary ADMET and PK models. The agent performs critical tasks such as parsing CRO reports, quality-checking dose-response curves, and managing synthetic resource queues.

In comparative testing against Claude Opus 5 and GPT-5.6 Sol, Indy demonstrated a significant lead in accuracy, correctly interpreting 84 real-world dose-response curves where competitors faltered. According to Nicholas Perl, head of chemistry at Anvia Therapeutics, the assistant surfaces data patterns that would otherwise be missed, allowing teams to prioritize scientific strategy over administrative maintenance. By shifting the burden of routine analysis to AI, Inductive Bio intends to reduce the frequency of costly errors that can stall drug development programs for months.