Felicis leads $7.5M bet on Unsiloed AI to make enterprise documents usable by AI agents
Unsiloed AI raises $7.5âŻM led by Felicis to turn complex enterprise documents into structured data that AI agents can reliably use.

Introduction
Unsiloed AI, a SanâFranciscoâbased startup, announced a $7.5âŻmillion seed round led by Felicis Ventures, with participation from YâŻCombinator, Orange Collective, Uncorrelated Ventures, and angel investors AarthiâŻRamamurthy and BethâŻTurner. The funding will accelerate the companyâs mission to make the wealth of information locked inside tables, charts, and other complex document elements usable by AI agents across finance, legal, healthcare, and other enterprise sectors.
The problem with todayâs documentâAI pipelines
Enterprises rely on PDFs, spreadsheets, and scanned reports for daily decisionâmaking. While large language models (LLMs) excel at generating text, they struggle with the precise extraction of structured data from visually rich documents. Errors in interpreting a financial table or a legal clause can lead to costly downstream mistakes, especially in regulated industries such as banking, hospitals, and law firms. Existing solutionsâgeneralâpurpose AI models, cloudâbased document services, and niche startupsâoften treat the entire page uniformly, wasting compute on irrelevant sections and still producing hallucinations.
Unsiloedâs technical approach
Unsiloed tackles the challenge with a âdualâstreamâ architecture that processes text tokens and image tokens separately. A heatâmapâbased chunking algorithm first identifies highâvalue regionsâfigures, numbers, and complex layoutsâso that computational effort is focused where it matters most. A specialized decoder then incorporates industryâspecific terminology, helping the system distinguish between contextâdependent meanings (e.g., âinterestâ in finance vs. âinterestâ in legal contracts).
In a technical brief shared by the company, a test on 542 financial documents (about 84âŻ000 data points) reported 99.78âŻ% extraction accuracy, outperforming Microsoftâs Azure Document Intelligence and the customerâs inâhouse tool. Unsiloed also claimed a leading score of 88.0 on the public olmOCRâBench benchmark, compared with 66.0 for Reducto and 39.9 for Unstructured. While these figures have not been independently verified, they illustrate the startupâs focus on precision and costâefficiency.
Early traction and market positioning
Unsiloed is already piloting its technology with customers in finance, legal services, and healthcare. One unnamed bank uses the platform to process internal documents, while a law firm has deployed it onâpremises for contract and caseâfile analysis. VoiceâAI startup BolnaâŻAI also leverages Unsiloed to ingest knowledge bases for building conversational agents.
The market for documentâprocessing AI is crowded, featuring both specialized startups and major cloud providers. Unsiloed differentiates itself by combining visualâaware models with domainâaware decoding, and by pricing the service at roughly one cent per pageâa model that reflects its selective compute strategy.
Use of the new capital
With seven employees at the time of the announcement, Unsiloed plans to allocate the seed funding toward three core areas:
- Talent acquisition â hiring additional researchers and engineers to refine the dualâstream models and expand the proprietary dataset pipeline.
- Sales and goâtoâmarket â building a dedicated sales team to target enterprise accounts in regulated sectors.
- Data and product development â creating curated datasets that capture industryâspecific document formats, further improving accuracy and reducing hallucination risk.
Practical takeaways
- For enterprises: Unsiloed offers a way to unlock structured data from legacy PDFs and scanned reports without fully trusting blackâbox LLM outputs.
- For AI developers: The dualâstream and heatâmap techniques illustrate a promising direction for handling multimodal document content.
- For investors: The seed round signals confidence in niche AI infrastructure that addresses realâworld compliance and accuracy challenges.
--- Unsiloed AIâs $7.5âŻM seed round underscores a growing demand for reliable, costâeffective document extraction that can feed AI agents with trustworthy data. As enterprises continue to automate knowledge work, solutions that combine visual precision with domain awareness are likely to become essential components of the AI stack.
Source
Read original sourceWhy we picked this
Reports funding for an AI document-processing startup, relevant AI business news.