1910

1910 offers the ITO platform, an AI-native infrastructure designed to address declining productivity in pharmaceutical R&D, where developing a single drug can take over a decade with a low probability of success. The platform tackles barriers of data scarcity, model training, computing power, and wet lab validation by integrating massive multimodal data, frontier AI models, and high-throughput lab automation into a single iterative loop for small and large molecule drug discovery.

The ITO platform's core capabilities include a Model Foundry with hundreds of AI/ML models and a multi-AI agent system for multi-parameter optimization. It ingests data from wet lab proxy assays, computational simulations, and ground truth biological data, orchestrated through a data pipeline. The platform is cloud-based on Microsoft Azure, leveraging partnerships for multi-GPU compute, and supports federated learning for collaborative model training without sharing private data. Specific AI models mentioned include PEGASUS for macrocyclic peptides and CANDID-CNS for central nervous system penetration.

The platform is built for biopharma companies and research partners seeking to apply AI to drug discovery. It is offered as a cloud-based infrastructure, with partnership and consortium models available. The website does not specify free plans, trials, or seat-based pricing, but indicates the platform can be replicated across major cloud providers.