Sarthak
Jain

There is a specific kind of professional suffering that no one talks about at conferences: the accounts payable manager at a mid-sized manufacturer who spends their Tuesday extracting invoice numbers from PDFs and entering them into SAP by hand. Multiply that person by several million. Now you understand why Sarthak Jain founded Nanonets.

Jain arrived in San Francisco via IIT Gandhinagar and Bangalore, carrying a Director's Gold Medal, two startup exits in his head (Cubeit's acqui-hire by Myntra in 2016), and a very specific observation about the world's knowledge economy. "The internet was going to kill paper," he told investors, "but businesses today are producing more documents than ever, just in new forms. Email, PDF contracts, whitepapers, etc. There are millions of highly skilled professionals stuck looking for needles in haystacks."

That observation, made in a Y Combinator batch in winter 2017, became the thesis behind Nanonets. Not another productivity tool. Not another dashboard. A machine that reads - invoices, purchase orders, contracts, insurance claims, KYC documents, bills of lading - and routes the extracted data where it needs to go, without a human in the loop.

By the time Nanonets raised its $29 million Series B from Accel in March 2024, the company had quietly become infrastructure for enterprises that don't advertise what they use. More than a third of the Global Fortune 500 was running Nanonets somewhere in their back office - finance, accounting, operations, procurement. The kind of work that doesn't make it onto keynote slides but decides whether a company's cash flow reporting is accurate or not.

Then came the growth numbers that made people look twice: $13.1M ARR in 2023, $40M in 2024, $100M by September 2025. A 150% year-on-year jump. With 130 people. In San Francisco.

At IIT Gandhinagar, Sarthak Jain was the kind of student who organized the annual cultural festival (Blithchron), helped build the basketball court for the Inter-IIT sports meet, and graduated with a B.Tech in Electrical Engineering with a Minor in Computer Science - earning the Director's Gold Medal for outstanding overall performance and a scholarship for excellence in social work and leadership.

And then he turned down his campus placement offer.

In 2014, Jain co-founded Cubeit with his IIT Gandhinagar batchmate Prathamesh Juvatkar and a third co-founder, Nithinkumar Gadiparthi. Cubeit was a content aggregation app - think bookmarks and web-clipping for mobile, useful for indoor locations like malls and museums. Accel Partners and Helion Venture Partners put in $3 million in seed funding. By 2016, Myntra (Flipkart's fashion portal) had acqui-hired the whole team.

The acquisition gave Jain two things: runway to observe how a major Indian e-commerce company handled its product operations, and a co-founder relationship with Juvatkar that had been stress-tested across two years of building. As Senior Product Manager at Myntra, he ran Feed and Engagement - learning the enterprise product muscle he'd need for what came next.

What came next was Nanonets. Founded in 2017, same year as the Myntra exit. The pattern - exit, reflect, build - compressed into months.

The elevator pitch has evolved. In 2017, Nanonets was OCR-plus-AI: point it at a document, get structured data back. Custom models, no-code, accessible to businesses that didn't have machine learning engineers. The OCR market was large and underserved at the enterprise level.

By 2024, the company had quietly repositioned. The $29M Series B announcement centered on "autonomous AI agents for business workflows" - not just reading documents, but acting on them. Routing an invoice through approval workflows. Reconciling accounts payable against purchase orders. Screening tenants. Processing insurance claims. KYC verification. The complete back-office loop, end to end.

The technology stack underneath - PyTorch, TensorFlow, React, TypeScript, Python, Kubernetes, PostgreSQL on AWS and Google Cloud - is what you'd expect from a company that has been building ML infrastructure since before large language models made every startup an "AI company." Nanonets was doing commercial machine learning before it was fashionable. Jain has spent 15+ years in the field, working on ML applications before most VC-backed AI startups existed.

The G2 Fall 2024 report recognized Nanonets as the Leader in the global OCR market, handing the company 12+ badges. But Jain's ambition runs further than OCR market share. The pivot toward autonomous AI agents positions Nanonets closer to the accounts payable automation and workflow orchestration markets - categories that include Automation Anywhere, UiPath, and Appian, but from a native-AI approach rather than a robotic process automation legacy.

In September 2025, Jain joined the AI Agents Summit panel on "Open Source vs. Closed Source AI Agents" - a signal of where the company's R&D attention sits. The debate is material: enterprise customers want control, auditability, and on-premise options; the best models often sit behind proprietary walls. Nanonets navigates both worlds.

Jain's advice on when to start a company cuts against most of what gets shared at founder meetups: "Starting in college is a bad idea generally. You are better working for someone else unless you have a strong why." This from a person who turned down campus placements, sold his first company by 27, and built a second one to $100M by his mid-thirties. The nuance matters.

He writes. Not just the 66+ technical blog posts on the Nanonets site - guides to data extraction, LayoutLM, intelligent document capture, purchase order automation. He maintains a Substack newsletter at sarthakjain.com focused on "Products, Startups and Products made by Startups." It's not ghostwritten. The prose is direct, practical, occasionally contrarian.

He plays table tennis and basketball. The basketball court at IIT Gandhinagar's campus - the one built for the Inter-IIT sports tournament - was partly his project as a student. The same person who built institutional infrastructure at 21 is now building infrastructure that processes Fortune 500 financial documents.

He credits IIT Gandhinagar's faculty for his entrepreneurial development more than he credits the investor ecosystem. Which is the kind of thing that's either genuine or performed, and everything in Jain's track record suggests the former.