A researcher forms a hypothesis. Someone still has to prepare the experiment, run the procedure and work out what the result changes. [1]
AI can help with papers, analysis and proposed experiments. The next step still takes place in the laboratory. [1]
Primordiant is building robotic experimentalists for that step.
The company calls its approach Embodied Scientific Intelligence: scientific reasoning and physical action within the same system. The aim is to observe, form a hypothesis, experiment and learn from the result. [1][2]
The experimentalist is being developed with contact with the physical world at its center. [1]
A scientific question that can change
A fixed protocol can be repeated many times. Discovery also changes the protocol: a surprising result can call for a different procedure or a new instrument. [1]
Primordiant’s goal is a general-purpose experimentalist that can adapt as the question changes. The system is being designed to work within existing laboratory processes. [1]
Scientific reasoning determines what to test. Physical action carries out the test. Observations from the experiment then inform the next hypothesis. [1][2]
The company is developing the reasoning and physical experimentation together, so an investigation can change course as results arrive. [1]
Physics and robotics on the same founding team
Owen Tower and Cooper Niu both earned physics PhDs at Brown. Tower’s work includes computational methods, microfluidics and experimental physics, with research experience at NASA JPL and CERN. [2]
Niu’s research includes particle physics and an autonomous system that proposes and refines physics theories. Niu previously worked on neutrino research at Fermilab. [2]
Benedict Quartey earned a computer science PhD at Brown, researching capable, verifiable robots and spatial world models. Quartey also developed AI systems at the Boston Dynamics Robotics & AI Institute. [2]
The founders bring theoretical, computational and experimental research together with robotics. Primordiant was founded in 2026 and joined YC’s Fall 2026 batch. [2]
From proposing an experiment to carrying it out
Here is the research loop Primordiant is working toward.
A scientific system observes a result, proposes a hypothesis and carries out the experiment needed to test it. The result changes what it tries next. [1]
For laboratory researchers, the intended capability is an experimentalist that can adjust its physical work as the investigation develops. The company is inviting laboratories to register for early beta access. [1]
The starting point is collaboration with labs exploring how physical experimentation and AI reasoning can develop together. [1]
Where Primordiant ranks
Primordiant isn't in a numbered ranking yet. Products get a position once enough companies compete on the same task.
Current rankings. Positions come from public evidence, and no company can pay for a place. How we rank
The bigger picture
A scientific idea becomes more valuable when it can be tested and changed by the result. [1]
Primordiant is building toward a system that joins the reasoning and the experiment. [1][2]
Scientific intelligence, with a way to act in the world.
Written by SOTA2 from Primordiant’s public materials. Customer experiences and product plans are reported by Primordiant. Ranking analysis and comparison data are from SOTA2.