Sapient Intelligence launches PRAXIST Beta, an autonomous AI R&D system that independently tests and validates potential solutions to complex problems by defining goals, parameters, and budgets for autonomous experimentation, functioning as an R&D capacity multiplier.

PRAXIST Beta enables users to set goals, parameters, and budgets while the system conducts autonomous experiments and evaluations, expanding researchers’ capability to explore multiple hypotheses.

PRAXIST (Beta) is positioned as an autonomous AI R&D platform that explores and validates technical solutions for complex problems without replacing human expertise.

Key figures include William Chen (Co-Founder) and Jin Li (Chief Scientist), with a global team spanning Singapore, Palo Alto, and Beijing.

Sapient Intelligence builds on its HRM architecture to enable deeper reasoning and more efficient training, supported by a global team of over 50 researchers across Singapore, Palo Alto, and Beijing.

Internal evaluations show PRAXIST achieving 49 highest-level results across 75 Kaggle ML benchmarks with an estimated model cost of $3,000, versus Claude Code achieving 34 top results at about $38,000 under the same conditions.

Proof-of-concept capabilities include a rocket-landing simulation with 100% accuracy within 12 hours (TRL 3) and an industrial robotic SLAM improvement cutting error from 9.37 cm to 5.01 cm in three days.

Partner experiments report 100% rocket-landing success within 12 hours and SLAM error reduction to 5.01 cm in three days, signaling rapid, concrete capabilities.

Jin Li, Chief Scientist, frames autonomous discovery as the next frontier and positions PRAXIST as a stepping stone for organizations to pursue more experiments while keeping human expertise central.

PRAXIST acts as an R&D capacity multiplier, enabling organizations to augment teams with faster, broader research, testing multiple hypotheses and learning from both successes and failures.

Sapient Intelligence emphasizes PRAXIST as a long-horizon enabler that augments human teams rather than replacing them, promoting broader exploration of problems and solutions.

PRAXIST can operate in private or customer-controlled environments to protect proprietary data, with initial target sectors including AI, engineering, robotics, manufacturing, finance, health, and drug discovery, and plans to broaden to business optimization.