Meet PRAXIST: an autonomous AI lab assistant that runs R&D experiments for you

Singapore-based AI company Sapient Intelligence has launched the beta version of PRAXIST, an autonomous AI system designed specifically for research and development work. Instead of focusing on chat or content generation, PRAXIST aims at hard technical problems—buggy systems, performance bottlenecks, or complex design questions—and tries to push them forward like a tireless research assistant.

The workflow is roughly: engineers feed in their existing materials—research papers, technical specs, bug reports, or design docs. PRAXIST then proposes multiple lines of attack, automatically running code, simulations, or test cases where possible. As results come back, it prunes bad ideas, refines promising ones, and surfaces a ranked set of potential solutions in a dashboard that explains what was tried and under which conditions it worked.

Sapient frames this as a step beyond today’s coding copilots: the goal is not just to autocomplete functions, but to explore unfamiliar solution spaces and generate new hypotheses. For now, PRAXIST is available as a gated beta, with early access targeted at software and AI teams. The company says it ultimately wants to extend the platform to other R&D-heavy areas, from materials science to hardware design, where the ability to run many small experiments quickly can translate directly into faster innovation.