Participated · Seed · Feb 2026

Adaption Labs is an AI research company focused on moving beyond static, parameter-frozen models by creating systems that can learn continuously from real-world interactions. Co-founded by Sara Hooker and Sudip Roy, the team emphasizes efficiency and adaptability over brute-force scale, designing models that structurally understand user needs and adjust their behavior on-the-fly without changing core weights. This approach aims to deliver more responsive, resource-efficient AI capable of evolving alongside user requirements. The startup plans to channel recent capital into expanding operations and accelerating product development. Based in San Francisco, the company positions itself at the forefront of adaptive intelligence research, targeting applications where traditional large-scale models are too rigid or compute-heavy. Although no revenue or user figures were disclosed, the sizeable seed investment underscores investor confidence in the technology’s commercial potential.

Led · Seed · Nov 2025

Onepot AI is building an end-to-end platform that fuses advanced machine learning with highly automated laboratories to shorten chemical synthesis cycles from months to days. Its proprietary AI chemist, Phil, designs and optimizes reaction pathways while the company’s first physical lab, POT-1, executes and meticulously records every experimental variable. This closed-loop data capture produces reproducible, high-quality datasets that continuously improve the system’s predictive power, enabling the expansion of chemical space previously considered impractical. Biotech and pharmaceutical clients can order target compounds from Onepot’s catalog, after which the company synthesizes and ships the molecules. Founders Daniil Boiko and Andrei Tyrin claim their approach can at least double the pace of traditional drug discovery workflows. Fresh capital will fund a second laboratory in San Francisco, significantly increasing commercial capacity. Ultimately, Onepot AI aims to remove synthetic feasibility as a constraint on medicinal chemistry innovation.

Participated · Series B · Nov 2025

Gridware is a San-Francisco-based grid-technology company whose pole-mounted Gridscope devices capture mechanical, electrical, and environmental data to enable Active Grid Response (AGR). Tens of thousands of these sensors are already installed on U.S. distribution poles, and the firm’s utility partners collectively serve roughly 40% of American electricity customers. The system pinpoints the location and cause of faults, helping crews respond faster and, in some cases, avoid failures altogether. Early deployments include Duquesne Light Company, which rolled out 1,700 additional devices after a successful pilot and subsequently won a PUF innovation award. With the fresh capital, Gridware plans to expand manufacturing capacity, scale deployment operations, and hire across hardware, firmware, software, data science, and machine-learning teams. The company is also preparing for international pilots in the United Kingdom and continental Europe. Ultimately, Gridware aims to help utilities worldwide improve reliability without expensive system overhauls.

Participated · Seed · Nov 2025

Red Queen Bio is a newly launched company spun out of HelixNano that focuses on "scaling biological defenses" as artificial intelligence capabilities accelerate. The firm’s platform combines state-of-the-art AI models with lab automation and reinforcement learning to systematically map potential AI-enabled biothreats. Using those insights, the company intends to design and stockpile medical countermeasures in advance of possible outbreaks. The founders emphasize that the scientific work begins with the close of the seed financing, indicating the company is still pre-commercial. No revenue, user, or other operating metrics have been disclosed. The overall vision positions Red Queen Bio at the intersection of synthetic biology, machine learning, and public health preparedness.

Participated · Seed · Oct 2025

Electroflow is building a three-step, fully electric system that extracts lithium from subterranean brines and converts it directly into lithium-iron-phosphate powder for battery manufacturers. The company’s proprietary cell design absorbs lithium ions in one pass and releases them into a carbonate solution in the next, eliminating several traditional refining steps. Prototype testing on brines from a California geothermal site validated the chemistry, and a Version-1 unit due by year-end targets production costs of about $5,000 per metric ton of LFP. At commercial scale, management projects costs below $2,500 per ton—roughly 40% cheaper than current Chinese suppliers—while fitting the equipment inside a 20-foot shipping container capable of producing 100 tons of material annually. The process consumes minimal electricity (roughly equal to a single U.S. household for 50 tons of output) and largely recycles its water. By leveraging abundant domestic brines, Electroflow aims to lower electric-vehicle battery prices by up to 20% and build a U.S. supply chain independent of Chinese incumbents. The company is led by co-founders Eric McShane (CEO) and Evan Gardner, both veterans of battery-materials research.