California-based Aureka Biotechnologies has closed a USD 100 million Series B financing round to advance its AI-native drug discovery platform and internal antibody pipeline, bringing total capital raised to nearly USD 200 million since the company's founding in 2023.
Granite Asia funded the first tranche exclusively, while a prominent but unnamed strategic investor led a subsequent tranche. HighLight Capital (HLC) participated alongside follow-on investments from existing shareholders MPCi and NRL Capital. Proceeds will be directed primarily toward large-scale training of next-generation biological foundation models and upgrading the company's Lab-in-the-Loop infrastructure — a closed-loop system integrating AI agents, high-throughput single-cell functional screening, and automated wet-lab validation. The Series B follows a USD 35 million Series A+ completed in April 2026, which was led by HSG and included MPCi, BioTrack Capital, and several returning investors. Prior to that, a Series A co-led by 5Y Capital and Qiming Venture Partners closed in November 2025.
Aureka is a direct spinout from the University of California, Irvine, where co-founders Weian Zhao, who serves as chief executive officer, and Alon Wellner, now chief innovation officer, developed the core platform technology. The company's wet-lab engine centres on Autonomous Hypermutation Yeast Surface Display (AHEAD), a continuous antibody evolution system described in a 2021 Nature Chemical Biology publication, which mimics somatic hypermutation by coupling an error-prone orthogonal DNA replication system to yeast surface display. This generates large-scale, functionally annotated proprietary data that feeds back into Aureka's AuraIDE biological foundation model — trained on internal protein co-evolution data and reported to rank among leaders on protein structure prediction and de novo molecular design benchmarks, including the FoldBench v1 antibody-antigen structure prediction evaluation. An open-source version, OpenDDE, has been made publicly available.