Basecamp Research

Astromech is a Delaware-incorporated startup with offices in Austin, Texas, spun out of de-extinction company Colossal Biosciences in 2025 by entrepreneur Ben Lamm and Harvard geneticist George Church. It is building what it calls a Large Life Model, a predictive system for biology trained on multi-species functional genomic data accumulated at Colossal. The platform combines transcriptomic data, cross-species comparison and ancestral state reconstruction to answer three questions: where a living system is heading, where it is most likely to break, and which regulatory mechanisms drive the change. Two model engines work in parallel, one applying deep learning to find patterns across species, the other reverse-engineering regulatory logic along evolutionary history. By focusing on ancestral regulatory states rather than protein sequence alone, the company argues it can model complex traits such as morphology and lifespan. Its first model targets aging and longevity, mapping 46 longevity-associated genes across species including the Asian elephant and the bowhead whale. The company is still in build-out mode, with 28 roles listed publicly.

Network Bio is a Palo Alto, California biotechnology company founded in 2022 that builds disease-specific AI models from human biological data. Its vertically integrated platform combines a biological research network that sources, links and structures tissue, blood and longitudinal clinical outcomes from leading U.S. academic medical centers (Mass General Brigham, University of Pennsylvania, Duke University, University of Colorado Anschutz and UCSF, spanning 500,000+ patients across oncology, immunology, metabolic and cardiovascular disease) into AI-ready datasets, with a bio-native AI architecture purpose-built to learn from multimodal biological data, overcome technical confounders and batch effects, and generate interpretable representations of disease biology that generalize across diseases, tissues and modalities.

Ecotone (Ecotone AI) is an AI genomics company based in San Francisco, with a team in Brooklyn, New York, founded in 2023 by eMalick G. Njie. It trains genome foundation models that read the human genome as a 'first language'; its flagship model dnaSORA is a diffusion transformer combining generative and discriminative capabilities to pinpoint disease-causing sequences among roughly 3 billion bases, which the company says can surface disease mechanisms from dozens rather than thousands of patient samples. It targets pharma and biotech drug developers working on rare genetic diseases and aims to cut genome-interpretation costs.

Enveda Biosciences, founded in 2019 in Boulder, USA, analyzes molecules produced by living organisms to build a large chemical information database for developing treatments for diseases such as atopic dermatitis and asthma.

Rivercell, founded in Paris in summer 2025, is building an AI 'world model of the cell': an automated wet lab generates large-scale multimodal perturbational single-cell data to train an AI Virtual Cell model that predicts how human cells respond to drugs and genetic changes.

Tahoe Therapeutics, founded in 2021 in South San Francisco, develops AI models of human cells trained on large-scale single-cell datasets for drug discovery.