# Praxist

> Praxist is an autonomous research product for measurable research and engineering problems. It is developed by Sapient Intelligence Pte Ltd.

Praxist turns an already runnable project into a continuous, evidence-driven research loop. Parallel research agents develop candidate solutions, evaluators convert results into structured evidence, and a PI/Chair planning panel uses that evidence to direct the next generation.

## Canonical entities

- **Product:** Praxist
- **Company:** Sapient Intelligence Pte Ltd
- **Category:** Autonomous research system and research orchestration software
- **Product release date:** 28 August 2026
- **Official website:** https://praxist.sapient.inc/en
- **Documentation:** https://praxist.sapient.inc/en/docs
- **Source repository:** https://github.com/sapientinc/PRAXIST
- **Contact:** https://praxist.sapient.inc/en/contact

## What Praxist requires

- A project that already runs in the intended environment.
- A measurable objective with a clear optimization direction.
- A repeatable evaluator, simulator, benchmark, or scoring path.
- Human review for deployment and production decisions.

## What Praxist produces

- Candidate implementations and experiment artifacts.
- Evaluator-backed metrics and structured findings.
- A durable frontier of verified evidence.
- A complete lineage showing how results were produced.
- A research agenda for the next generation.

## Primary resources

- [Product overview](https://praxist.sapient.inc/en)
- [About](https://praxist.sapient.inc/en/about)
- [Measured examples](https://praxist.sapient.inc/en/examples)
- [Technical documentation](https://praxist.sapient.inc/en/docs)
- [Full machine-readable corpus](https://praxist.sapient.inc/llms-full.txt)

## Boundaries

Praxist is not a replacement for missing datasets, simulators, baselines, evaluators, safety review, or production approval. It is intended for research paths whose outcomes can be executed and measured on a computer.