Discovery Loop automates the experimental loop of scientific discovery by building AI systems that propose, run, and analyze thousands of experiments in parallel. The company starts with machine learning research and engineering, using these automated capabilities to optimize its own stack before expanding into other science and engineering domains. Its ultimate goal is to tackle NAE Grand Challenges like better medicines and economical solar energy.
- Artificial Intelligence
- AI Agents
- Software Only
Funding
Funding not disclosed
Founders
Product
Problem
Scientific progress relies on sequential, manual experimental loops—proposing an experiment, implementing it, running it, examining results, and iterating—which are slow and labor-intensive. This manual approach limits the scale and speed of discovery across science and engineering domains, bottlenecking innovation.
Solution
Discovery Loop automates the entire experimental loop, enabling the parallel execution of thousands of experiments and drastically compressing iteration time. The company initially focuses on automating machine learning research and engineering, using these capabilities to rapidly optimize its own technology stack before expanding to other domains. Its AI systems are designed to automatically solve important problems in machine learning, science, and engineering, and ultimately take on National Academy of Engineering Grand Challenges such as engineering better medicines, advancing health informatics, making solar energy economical, providing access to clean water, and securing cyberspace.
Target Audience
The primary audience is research teams and engineering organizations in machine learning, science, and engineering domains that require faster iteration cycles and scalable experimentation to accelerate discovery.
Features
- AI-driven experimental loop automation that proposes, implements, runs, and analyzes experiments without human intervention
- Parallel execution of thousands of experiments to compress iteration timelines and increase output quality
- Self-optimizing technology stack, using automated ML to improve its own systems before external deployment
- Domain-agnostic design intended to handle any learning loop with measurable outcomes in science and engineering
- Strategic alignment with NAE Grand Challenges to target high-impact societal problems in medicine, energy, water, and cybersecurity