Retab, a French AI startup, has secured $3.5 million in pre-seed funding concurrent with the debut of its platform.
This funding round was supported by prominent early-stage investment firms like VentureFriends, Kima Ventures, and K5 Global, as well as Eric Schmidt through StemAI, Olivier Pomel, CEO of Datadog, and Florian Douetteau, CEO of Dataiku.
“Retab is the OS for reliably extracting structured data,” said Louis de Benoist, co-founder and CEO of Retab. “It wraps the best models in a layer of logic that actually makes them usable with error handling and structured outputs. That’s what developers need if they want to build production apps, not just prototypes.”
This investment will aid in the development of the platform and the growth of its community, as the company expands its infrastructure to accommodate increasing demand from vertical AI startups and internal innovation teams, the company said.
Retab operates as a developer platform and SDK that transforms document processing in the era of large language models. Developers specify the data schema they require, and Retab manages everything else from dataset labelling and evaluations to automated prompt engineering and model selection.
The platform guarantees high performance through intelligent checks and balances. It uses self-optimising schemas where an AI agent refines instructions based on user documents for maximum accuracy before deployment.
Its model-agnostic routing directs tasks to the best-performing model, prioritising cost, speed, or accuracy, potentially cutting expenses by up to 100 times. Additionally, guided reasoning and a k-LLM consensus mechanism ensure reliable results by requiring models to think step-by-step and assess uncertainty collaboratively.
Retab is expanding its platform to apply reliable data extraction methods to websites and is launching integrations with automation platforms like n8n, Zapier, and Dify. The company aims to become an intelligent middleware layer between unstructured data and AI agents.
Florian Douetteau, co-founder and CEO of Dataiku and investor in Retab, said, “The AI-fication of the economy depends on the capability to convert operations based on millions of documents into verified, structured data that autonomous systems can utilise. On a large scale, this process hinges on quality control, cost efficiency, and rapid implementation.”
