Retab has launched from stealth with a pre-seed funding round totalling USD $3.5 million to support its document AI platform and developer-first tools.
The company's funding round was supported by several early-stage investors, including VentureFriends, Kima Ventures, and K5 Global, alongside individual investors Eric Schmidt (via StemAI), Olivier Pomel (CEO of Datadog), and Florian Douetteau (CEO of Dataiku). The capital is earmarked for further platform development and community growth as Retab looks to serve demand from vertical AI startups and innovation teams within enterprises.
Platform focus
Retab offers a software development kit (SDK) designed to help developers automate document processing tasks that use large language models (LLMs). The platform allows developers to define the data schema needed for their applications, while Retab manages data set labelling, evaluations, prompt engineering, and model selection. This orchestration is aimed at streamlining the document automation workflow and addressing common production challenges within the space.
Louis de Benoist, Co-Founder and Chief Executive Officer, said, "People keep building demos that look like magic, but break the moment you put them into production. We lived that pain ourselves. Wiring up fragile pipelines just to extract a few fields from a PDF. We built Retab because it's the developer-first platform we always wished we had."
Retab's founding team, composed of engineers with backgrounds in logistics workflow automation, identified greater value in the orchestration layer they had previously developed to make AI models usable at scale. This tooling now forms the basis of the Retab platform, which is being utilised by multiple companies to extract structured data from a variety of document types.
System capabilities
The platform is positioned not as an LLM itself, but as a middleware solution that integrates with LLMs from major providers such as OpenAI, Google, and Anthropic. Retab's system ensures accuracy and reliability by enabling developers to specify required outputs, while the platform oversees processes to maintain verifiable data extraction quality.
Key features of the platform include self-optimising schemas, which use AI agents to test and refine extraction instructions before deployment. The model-agnostic system benchmarks and routes each automation task to the most suitable LLM for requirements such as cost, speed, or accuracy. Retab states this approach can reduce costs by up to 100 times compared to conventional solutions.
Another element is guided reasoning and k-LLM consensus, where multiple models are used in parallel to reach a quantified consensus, supporting error reduction and increased trustworthiness in outputs.
"Retab is the OS for reliably extracting structured data. It wraps the best models in a layer of logic that actually makes them usable with error handling and structured outputs. That's what devs need if they want to build production apps, not just prototypes," said de Benoist.
Industry adoption
Retab reports adoption across logistics, finance, and healthcare sectors. A logistics firm has employed Retab to find the smallest, fastest LLM configuration for 99% accuracy, reducing operational costs. A financial services client utilises the platform to extract both quantitative metrics and qualitative risk factors from long quarterly reports, a process that previously required several days of manual analysis. Other customers are using the system to automate claims processing, medical records, identity verification, and onboarding processes.
Florian Douetteau, Co-Founder and Chief Executive Officer of Dataiku, and investor in Retab, observed, "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 utilize. On a large scale, this process hinges on quality control, cost efficiency, and rapid implementation. The team at Retab understands this thoroughly and is uniquely positioned to solve it for the thousands of AI first companies that are emerging."
Outlook
The company is extending its extraction capabilities to handle data from websites and is launching integrations with workflow automation platforms such as n8n, Zapier, and Dify. Retab's stated long-term strategy is to serve as intelligent middleware between unstructured data and AI systems, with the objective to make such data usable and programmable for a range of applications, from contract processing to customs management.
With a workforce of ten and a growing developer user base, Retab aims to become a key component in AI infrastructure for document automation workflows.
