BASALT has built a SaaS platform that structures the engineering of AI agents around a continuous loop of experimentation, evaluation and production monitoring. The software lets technical teams test prompt variants, compare models, analyse edge-cases and document results while giving business experts a workspace to define quality criteria and validate scenarios. By centralising this workflow, the tool helps companies avoid regressions, improve agent reliability and surface production issues that only appear in real-world contexts. A second pillar of the product is fine-grained monitoring, which detects problematic responses in production and automatically feeds them back into new test scenarios. Founded by Guillaume Marquis and François de Fitte, the company is headquartered in San Francisco with its technical team in Paris. The newly secured capital will be used to accelerate product development and support hiring in both France and the United States. No revenue or user figures were disclosed in the articles.

Total raised

$5M

Funding rounds

1

Latest round

Equity

Latest activity

Dec 2025