Autoscience Raises $14M to Automate Machine Learning Research
Autoscience, a San Mateo-based AI lab, announced $14 million in seed funding to automate the research and development of machine learning models. General Catalyst led the round, with participation from Toyota Ventures, Perplexity Fund, MaC Ventures, and S32.
The company has built a virtual laboratory where AI systems function as researchers and engineers. These systems generate new algorithmic ideas, test them, and deploy validated models into production.
The bottleneck isn't compute or data
Machine learning teams face a specific constraint: human capacity to evaluate and implement new research. More than 2,000 machine learning papers publish weekly. No research team can assess every breakthrough while advancing their own work.
Autoscience's approach uses two core systems. Automated scientists generate and test new algorithmic hypotheses. Automated engineers optimize those validated discoveries and prepare them for deployment.
Early validation in competition
The system produced a peer-reviewed research paper accepted to the ICLR 2025 workshop, marking the first time an autonomous lab achieved this milestone. It also earned a Silver Medal in the Kaggle Santa 2025 competition, placing against 3,300 teams and becoming the first fully autonomous system to rank in a live, featured Kaggle competition.
Initial focus: high-stakes applications
Autoscience is targeting financial services, manufacturing, and fraud detection. The company positions its service as a replacement for in-house research divisions-delivering research output without headcount.
The $14 million will fund expansion to Fortune 500 and large private companies training specialized models. The capital also supports growth of Autoscience's engineering team.
Eliot Cowan, CEO, said: "We've reached a point where human intuition is no longer enough to navigate the complexity of algorithmic discovery. We aim to compress a decade of machine learning research into months."
Yuri Sagalov, Managing Director at General Catalyst, said the firm backed Autoscience because it addresses "an increasingly important challenge in machine learning: the pace and scalability of experimentation."
For researchers and data scientists evaluating tools that can accelerate model development, see AI Research Courses and Generative AI and LLM Courses.
Your membership also unlocks:
