Luminary Raises $22.0M

Graph AI

Graph AI is the pioneer of the industry's first AI-native Patient Safety Operating System (OS), engineered to unify the fragmented pharmacovigilance (PV) landscape into a single, cohesive infrastructure for both biopharma enterprises and Contract Research Organizations (CROs). Founded by enterprise technology veterans, the company replaces legacy, labor-intensive workflows with a deterministic, inspection-ready AI architecture that explicitly aligns with the FDA’s AI Credibility Assessment Framework, the EU AI Act's high-risk data governance controls, and the CIOMS Working Group XIV recommendations. Rather than acting as a traditional "black-box" point-solution, this Patient Safety OS delivers mathematical reproducibility and absolute data transparency across a comprehensive lifecycle portfolio: Graph Safety /intake for multi-channel ingestion; Graph Safety /nucleus for touchless case processing and multi-algorithm causality validation; Graph Safety /report for adaptive generative medical writing; Graph Safety /signal for predictive safety surveillance; and Graph Safety /assure & /comply for strict QPPV governance. From global biopharma companies and emerging biotechs to consumer healthcare, medical device manufacturers, and CROs, Graph AI combines deep industry expertise with agentic AI to help organizations streamline operations, uncover meaningful insights, strengthen compliance, and deliver measurable business outcomes.

Positron AI

Positron delivers vendor freedom and faster inference for both enterprises and research teams, by allowing them to use hardware and software explicitly designed from the ground up for generative and large language models (LLMs). Through lower power usage and drastically lower total cost of ownership (TCO), Positron enables you to run popular open source LLMs to serve multiple users at high token rates and long context lengths. Positron is also designing its own ASIC to expand from inference and fine tuning to also support training and other parallel compute workloads.