Unsiloed AI, an artificial intelligence startup founded by IIT Kharagpur alumni Aman Mishra and Adnan Abbas, has raised $7.5 million in seed funding in a round led by venture capital firm Felicis. The funding will help the company strengthen its document intelligence technology, expand its engineering team and scale its enterprise deployments.

The round also saw participation from Y Combinator, Orange Collective and Uncorrelated Ventures, alongside angel investors Aarthi Ramamurthy and Beth Turner.

The November 18, 2026, funding announcement comes as businesses increasingly explore AI agents to automate complex operations. Unsiloed AI is targeting a specific challenge in this transition: enabling AI systems to accurately understand and work with information stored in business documents.

Aman Mishra and Adnan Abbas Are Building the Future of Document Intelligence

Founded by Aman Mishra and Adnan Abbas, former batchmates at the Indian Institute of Technology (IIT) Kharagpur, Unsiloed AI is developing technology that converts complex documents into structured information that artificial intelligence systems can understand and process.

The company was part of Y Combinator's Fall 2025 batch. Its founding team brings experience from technology companies and organisations including Cognition, Snowflake and Stripe, as well as quantitative trading firms.

Headquartered in San Francisco, Unsiloed is positioning itself at the intersection of enterprise software, computer vision and AI infrastructure.

How Unsiloed AI Is Making Documents Readable for AI Agents

Most enterprise information does not exist in neatly organised databases. It is spread across contracts, financial statements, insurance records, healthcare documents, presentations, scanned files, tables and charts.

Traditional optical character recognition (OCR) technology can extract text from documents, but understanding the relationship between text, images, tables and visual layouts remains a more complicated task.

Unsiloed AI addresses this problem through computer vision and dual-stream vision models that process both textual and visual information. Its APIs allow software to read, extract, classify, split, edit and update information contained in complex documents.

The company's broader objective is to make documents as computable as databases, enabling AI agents to retrieve and modify information without losing the context in which it appears.

This capability could be particularly relevant to financial institutions, insurance companies, healthcare providers and law firms, where accuracy and document context are critical.

$7.5 Million Funding to Strengthen Enterprise AI Infrastructure

Unsiloed AI plans to deploy its newly raised capital across three key priorities:

  • Technology development: Improving its proprietary document models and advancing document understanding capabilities.

  • Engineering expansion: Growing its technical team to accelerate product development and research.

  • Enterprise deployments: Expanding cloud, private-cloud and on-premises infrastructure to serve organisations with different security and compliance requirements.

The startup also supports fully air-gapped deployments, allowing businesses to process sensitive documents within their own infrastructure rather than relying entirely on external cloud environments.

According to the company, its platform has already processed millions of pages for organisations ranging from early-stage AI startups to publicly traded companies and law firms.

Why Investors Are Paying Attention to Document AI

The next phase of enterprise AI will depend on more than powerful language models. AI agents must also access reliable information, understand its context and take actions based on what they find.

For industries such as banking, insurance, healthcare and legal services, documents remain central to everyday operations. An AI system that extracts incorrect figures from a financial statement or overlooks a clause in a contract could create significant operational risks.

This is where Unsiloed AI sees its opportunity.

By combining computer vision with document-processing models, the company aims to build an infrastructure layer between unstructured business information and AI agents.

Felicis' investment reflects growing investor interest in the infrastructure required to make enterprise AI more practical and dependable.

However, Unsiloed is entering a competitive market that includes specialist document-processing companies such as Reducto and Unstructured, alongside document intelligence offerings from major technology providers.

Its ability to differentiate through extraction accuracy, enterprise security, deployment flexibility and commercial adoption will be important as it scales.

What Unsiloed AI's Funding Means for the Indian Startup Ecosystem

Although Unsiloed AI is headquartered in the United States, its founders' IIT Kharagpur connection highlights the growing international presence of Indian engineering talent in advanced AI infrastructure.

The company's focus also offers an important lesson for emerging Indian founders: opportunities in AI are not limited to building general-purpose chatbots or consumer-facing applications.

Solving specialised enterprise problems, particularly in document processing, data extraction and workflow automation, represents another important area of opportunity.

With $7.5 million in fresh seed capital and backing from Felicis and Y Combinator, Aman Mishra and Adnan Abbas now have additional resources to take their document intelligence platform to a wider enterprise market.

The next milestone will be demonstrating that its technology can deliver reliable results at scale across industries where accuracy, privacy and operational trust matter most.

View all
Explore
The Unicorn Magazine

The Unicorn Magazine is a business and technology media platform & Magazine, decoding the ideas, people, and companies shaping the future of India. We go beyond headlines to uncover the stories, strategies, and insights behind startups, entrepreneurs, investors, and innovators.

More