Unsiloed AI Raises $7.5 Million to Make Documents AI-Ready By Elets News Network - 22 November 2026

Unsiloed AI co-founders Adnan Abbas and Aman Mishra posing in their office, one seated on a green armchair

Unsiloed AI, a document intelligence startup founded by two IIT Kharagpur graduates, has raised $7.5 million in seed funding led by Silicon Valley venture firm Felicis. The startup announced the round on 18 November 2026, builds vision models and APIs that turn complex enterprise documents into structured data that AI agents can read and act on.

Y Combinator, Orange Collective and Uncorrelated Ventures also joined the round, along with angel investors Aarthi Ramamurthy and Beth Turner, TechCrunch reported. Unsiloed AI said it will use the money to improve its document models, grow its engineering team and expand cloud and on-premises deployments for enterprise customers.


What Unsiloed AI builds for enterprise AI agents

Enterprises are rolling out AI agents for finance, cyber security and supply chain work, but much of the information those agents need sits inside PDFs, scanned files, tables and charts. General-purpose large language models often misread such layouts, and a single wrong figure can be costly for a bank, a hospital or a law firm.

Unsiloed AI breaks a document into its parts, extracts data from tables, charts and other elements, and returns it as structured output. Each extracted value is linked back to its exact place in the source file, so teams can check the result instead of trusting the model blindly.


  • Dual-stream vision models: text tokens and image tokens are processed in separate streams.
  • Heat map-based chunking: more compute is directed to figures, numbers and charts.
  • Domain-aware decoding: a specialised decoder handles industry terms that mean different things in different sectors.
  • Flexible deployment: private cloud, on-premises and fully air-gapped set-ups for sensitive documents.

“The problem with LLMs is they’re inherently probabilistic,” co-founder and CEO Aman Mishra told TechCrunch, adding that hallucinations in document extraction have “a massive negative impact downstream.”

Founders, customers and accuracy claims

Mishra co-founded the San Francisco-headquartered company with Adnan Abbas, his batchmate at IIT Kharagpur. The pair went through Y Combinator’s Fall 2025 batch, joining a growing list of Indian-founded AI startups backed by Y Combinator. Its founding team includes alumni of Cognition, Snowflake, Stripe and quantitative trading firms, Business Standard reported.

The company says it has processed millions of pages for clients ranging from early-stage AI startups to listed companies and law firms. Customers include a bank that processes documents with it and a law firm that runs it on its own servers for contracts and old case files. Voice AI startup Bolna AI uses Unsiloed AI to ingest knowledge bases for its voice agents.

In a test on 542 financial documents supplied by an enterprise customer, Unsiloed AI reported 99.78% accuracy across about 84,000 extracted data points, ahead of leading AI models, Microsoft’s Azure Document Intelligence and the customer’s in-house tool. The results have not been independently verified.

Why document AI matters for Indian enterprises

Unsiloed AI enters a crowded field that includes Reducto, Unstructured and offerings from Microsoft and Amazon. Its pitch is accuracy with traceability, which matters most in regulated sectors such as banking, insurance and healthcare, where Indian teams are already testing small language models for KYC tasks. A recent TCS enterprise AI study found that only 10% of firms let AI agents act alone, a gap that verifiable data pipelines aim to close.

“For AI agents to transform how enterprises work, they first need to understand the documents those businesses run on,” said Astasia Myers, senior venture partner at Felicis.

Key takeaways

  • Unsiloed AI raised $7.5 million in seed funding led by Felicis, with Y Combinator participating.
  • Its vision models turn complex documents into structured, traceable data for AI agents.
  • The founders, Aman Mishra and Adnan Abbas, are IIT Kharagpur batchmates.
  • Funds will go into model research, hiring and enterprise cloud and on-premises deployments.

Be a part of Elets Collaborative Initiatives. Join Us for Upcoming Events and explore business opportunities. Like us on Facebook , connect with us on LinkedIn and follow us on Twitter.

"Exciting news! Elets technomedia is now on WhatsApp Channels Subscribe today by clicking the link and stay updated with the latest insights!" Click here!

Related artificial intelligence