Retab Unveils Powerful Document AI Platform with $3.5M Pre-seed Funding

Retab, a San Francisco-based startup, has raised $3.5 million in pre-seed funding to launch its developer-first document automation platform. Backed by prominent investors, including Eric Schmidt and executives from Dataiku and Datadog, the platform promises to transform document processing. The platform uses large language models and AI agents to streamline document extraction with accuracy, reliability, and efficiency for industries such as logistics, finance, and healthcare.

In the rapidly advancing field of AI, document processing has remained one of the most challenging problems. As businesses around the world generate massive amounts of unstructured data, the need for solutions that can extract and organise this information with accuracy and efficiency has become critical. Retab, a new player in the AI landscape, is addressing this gap by launching a powerful document automation platform designed to streamline and optimise the way businesses handle documents.

Founded by engineers who were frustrated by the limitations of existing tools, Retab has emerged with a unique approach to document AI. Today, the startup announces its $3.5 million in pre-seed funding, backed by leading early-stage investors and high-profile individuals, including Eric Schmidt (via StemAI), Olivier Pomel (CEO, Datadog), and Florian Douetteau (CEO, Dataiku). 

This investment will fuel the growth of the platform, aimed at providing developers with an intelligent infrastructure for transforming unstructured data into clean, structured information at scale.

A Developer-First Document Automation Platform

Retab is built for developers who need reliable document processing in industries where document workflows are central. The platform eliminates the need for complex, fragile pipelines often found in document AI projects.

It empowers developers by simplifying the document extraction process, allowing them to define the schema of data they need, while the platform handles the rest – from dataset labelling and model evaluation to automated prompt engineering and model selection.

The platform is designed to work seamlessly with models from leading AI providers like OpenAI, Google, and Anthropic, but its real innovation lies in its ability to bridge the gap between raw document data and actionable, structured information. 

As Louis de Benoist, co-founder and CEO of Retab, explains, "People keep building demos that look like magic but break the moment you put them into production. We lived that pain ourselves. We built Retab because it’s the developer-first platform we always wished we had."

Addressing Industry Challenges

The problem of inefficient and inaccurate document processing is widespread across several industries, including logistics, finance, and healthcare. Retab’s platform is already being used by companies to automate time-consuming, error-prone tasks. 

For instance, a major trucking company leveraged Retab to find the smallest, fastest model configuration that could meet their 99% accuracy threshold, significantly reducing operational costs. Similarly, a financial services firm has been able to automate the extraction of quantitative metrics and qualitative risk factors from large quarterly reports, cutting down what used to be a days-long task into an efficient, automated process.

Retab's approach to document AI also ensures that the system remains flexible and adaptive. The platform incorporates a system of intelligent checks and balances that guarantees performance and accuracy. 

Key features include self-optimising schemas, which automatically test and refine instructions based on user data, and intelligent model routing, which benchmarks tasks and directs them to the best-performing model depending on the priority, whether it's cost, speed, or accuracy. 

This capability makes Retab a game-changer, providing cost savings of up to 100x compared to traditional methods.

Enhancing AI Integration with Safety and Trust

The need for safety and reliability in AI workflows is paramount, and Retab ensures that its platform provides just that. One of its key features is guided reasoning, which forces models to "think" step-by-step, allowing for clearer insights into decision-making processes. In addition, the platform uses a consensus mechanism between multiple models to quantify uncertainty, ensuring that results are trustworthy and verifiable.

As Florian Douetteau, co-founder and CEO of Dataiku, highlights, "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. The team at Retab understands this thoroughly and is uniquely positioned to solve it for the thousands of AI-first companies that are emerging."

Scaling for the Future

Looking ahead, Retab has ambitious plans for expansion. The company is developing the platform to handle more types of unstructured data, including information from websites, and is working on integrations with automation platforms like n8n, Zapier, and Dify. 

The long-term vision for Retab is to serve as the middleware layer between unstructured data and AI systems, enabling industries such as finance, healthcare, and logistics to efficiently leverage the power of AI without the complexity and fragility of traditional document AI pipelines.

Despite being a young company with only ten employees, Retab is already well on its way to becoming a foundational player in the AI infrastructure space, providing businesses with the tools they need to automate document-heavy workflows and unlock the full potential of their data.

About Retab

Founded by engineers from Cambridge and École Polytechnique — Louis de Benoist, Sacha Ichbiah, and Victor Plaisance, Retab is an AI-driven platform that revolutionises document extraction. 

The all-in-one platform turns messy PDFs, handwritten scans, and other unstructured documents into clean, structured data. It automatically routes tasks to the best model for the job and adapts as better models emerge. Retab’s mission is to help teams replace manual work with fast, accurate, and self-improving workflows. 

Whether it's processing contracts, invoices, or compliance documents, Retab delivers reliable, scalable document automation solutions for enterprises across industries.