Rocket Vector develops causal AI products that help teams connect models, evidence, uncertainty, and practical next steps. Its focus is decision bottlenecks where predictions, research findings, and expert judgment are scattered, making it difficult to decide what to test, change, pursue, or stop. The company’s products are designed as decision-support layers that can complement existing tools rather than replace an entire technology stack.

Darkstar is a hosted workbench for building, inspecting, learning, and sharing Bayesian networks. Users can enter evidence, run inference, explore interventions and counterfactuals, analyze uncertainty, and move validated models toward runtime use. Darkstar also offers model-learning workflows, sharing features, and options for commercial SDK access and private deployment.

Navigator is a bundle-driven workspace for reviewing evidence from LeadScope drug discovery runs. It helps teams compare candidates, inspect supporting artifacts and citations, explore structured reports and charts, and consider actions such as pursuing, rescuing, modifying, or stopping a candidate. Navigator is approaching release, with preview access available by request.