Scientific research is drowning in data but short on time. Laboratories generate vast datasets and face an ever-growing firehose of new papers. However, much of the day is still spent on manual literature reviews, fragmented tools, and trial-and-error experimental design. Edison Scientific wants to change that by giving researchers AI “co-scientists” who can read literature, analyze data, and help plan experiments.
The San Francisco-based startup has raised a $70 million seed round at a valuation of approximately $250 million, jointly led by Spark Capital, Triatomic Capital, and leading US biotech investors, with participation from angels including Google chief scientist Jeff Dean and CrowdStrike co-founder Dmitri Alperovitch.
Give scientists research partners
Edison spun out of FutureHouse, a nonprofit AI biology institute founded in 2023 to train “AI scientists” for basic research. CEO and co-founder Sam Rodriques is a physicist and bioengineer known for his work in brain mapping and transcriptomics. In contrast, co-founder Andrew White has led several AI projects for chemistry, including ChemCrow and PaperQA.
At the core of Edison's platform is Cosmos, an “AI scientist” that can run long-term research campaigns on a single question. Given a goal and one or more datasets, Kosmos iterates through data analysis, literature search, hypothesis generation, and returns a fully cited research report. Some users say this compresses months of work into one day.
At Kosmos, specialized agents handle tasks such as literature synthesis, molecular design, and prior research. All of these tasks are accessed through a unified environment that integrates with existing laboratory data and workflows. Early projects include identifying therapeutic targets, mapping complex molecular structures, and narrowing down promising candidate molecules to shorten the path from idea to experimental execution.
Unlike Periodic Labs, Lila Science, and Chai Discovery, Edison doesn't own the entire stack, but rather helps scientists decide what to do next.
What's next?
Edison, which currently has about 30 employees, plans to use the new funding to hire in AI, biology, chemistry and engineering, and to turn early momentum into broader commercial expansion.
The roadmap includes increasing Kosmos' autonomy, deepening its integration with laboratory information systems and collaboration tools, and expanding its “assistant-style” agents to handle specific tasks such as data analysis and proactive search.
