They built India's national data platform. Then the tool they made for themselves became a $750 million company.
Prukalpa Sankar and Varun Banka spent seven years solving India's data problem at the national scale — building platforms for the PM's office, the UN, and NITI Aayog. The data chaos of doing that work forced them to build their own tool. That tool became Atlan. $206M raised. $750M valuation. Mastercard, Nasdaq, General Motors, and the NHS trust it with their most critical data.
Prukalpa Sankar & Varun Banka
Co-Founders · Atlan
A Bus Ride That Changed Indian Data
Prukalpa Sankar was about to graduate from Nanyang Technological University in 2012, with internship offers from ExxonMobil and Goldman Sachs already in hand. On a bus to campus, she met Varun Banka — same batch, same obsession with startups, same desire to do something that mattered.
At the Microsoft Imagine Cup that April, they ideated SocialCops: a big data platform for development decisions. They finished runners-up and returned to India to build it for real. Over the next six years, SocialCops built the national data platform used by India's Prime Minister's office, NITI Aayog, and the United Nations — helping route everything from rural healthcare logistics to agricultural insurance data.
"On the outside, you think 'cool projects,'" Sankar told Datanami. "But on the inside, it was not. It was just chaos every single day." Managing data across 200+ simultaneous projects, with every team working in different tools and formats, was ungovernable. They tried three separate data catalog and governance tools. None worked. The fourth time around — the one they built themselves — finally did.
The Internal Tool That Became a $750M Company
By 2019, Sankar and Banka faced a question most founders never get to ask: the side project built to manage their real company was more valuable than the real company.
Atlan — the internal data collaboration tool — had been stress-tested across 200 projects, dozens of data sources, and a team of analysts, scientists, and engineers who each worked in their own preferred tools. It had become, as Sankar describes it, a "GitHub for data teams" — a place where data has lineage, owners, definitions, and trust scores, connected across every system in an organisation.
They spun it out in 2019. Within two years, Forrester named it a Wave Leader for Enterprise Data Catalogs. Within three, it was the sole leader in G2's Active Metadata Management category — a category it effectively invented. By May 2024, GIC (Singapore's sovereign wealth fund) and Meritech Capital led a $105M Series C at a $750M valuation. Revenue had grown 7x in two years. Enterprise sales had grown 400% in a single quarter.
Context Is King — The AI Inflection Point
Every major enterprise AI initiative runs into the same wall. The model is fine. The problem is that the model knows nothing about the business it's being asked to reason about. What does "revenue" mean here — gross or net? Which Snowflake table has the authoritative number? Whose definition of "active customer" should the dashboard use?
Atlan's entire product is the answer to that question. Its Enterprise Data Graph — built by connecting 80+ data tools including Snowflake, Databricks, BigQuery, Salesforce, BI platforms, and AI LLMs — holds the institutional knowledge that makes AI agents actually useful in production.
"In 1996, Bill Gates wrote that content is king," Varun Banka noted in 2025. "In 2025, context is king." The launch of Atlan's MCP Server — which pipes its enterprise context directly into any AI agent — completed the circle. The company that started on a bus in Singapore, building a data platform for India's poorest districts, is now the context layer that makes enterprise AI work for Mastercard, General Motors, and the NHS.
"The main hurdle isn't AI models — it's the lack of AI-ready data. Data enriched with business context, trust, and security. That's what Atlan builds."
— Prukalpa Sankar, Co-Founder & Co-CEO, Atlan
Company Timeline
- 2012
Prukalpa Sankar and Varun Banka meet at Nanyang Technological University, Singapore. They co-found SocialCops at the Microsoft Imagine Cup — runners-up. Mission: use big data for social good.
- 2012–18
SocialCops builds India's national data platform. Used by the PM's office, NITI Aayog, and the UN. Across 200+ data projects, the internal data collaboration tool they build proves more powerful than the projects it supports.
- 2019
Atlan spun out of SocialCops as a standalone company. First external funding from WaterBridge Ventures and Rajan Anandan. The product repositions from internal tool to enterprise data catalog platform.
- 2021–22
Named a Forrester Wave Leader for Enterprise Data Catalogs and G2 Leader in five categories — the sole leader in Active Metadata Management since the category's inception. Insight Partners and Peak XV back the company.
- Mar 2024
$27.5M funding led by Salesforce Ventures. Salesforce becomes both an investor and a key integration partner — signalling that Atlan is becoming foundational enterprise infrastructure.
- May 2024
$105M Series C led by GIC (Singapore's sovereign wealth fund) and Meritech Capital. Valuation reaches $750M. Revenue 7x over two years. 400% enterprise sales growth in Q1 2024. Total raised: $206M.
- 2025–26
Named Leader in Gartner Magic Quadrant for Metadata Management Solutions (2025) and Data & Analytics Governance (2026). Atlan MCP Server launched — bringing its context layer directly into every AI agent workflow.
Frequently Asked Questions
Who are the founders of Atlan and where did they study?
Atlan was co-founded by Prukalpa Sankar and Varun Banka, who met at Nanyang Technological University (NTU), Singapore in 2012. Both serve as co-CEOs. Before Atlan, they co-founded SocialCops — which built India's national data platform used by the PM's office and the UN.
What is Atlan's data catalog and how does it work?
Atlan scans an organisation's entire data ecosystem — connecting to 80+ tools including Snowflake, Databricks, BigQuery, Looker, Salesforce, and AI models — and creates a unified, living map of every data asset, its lineage, ownership, definitions, quality scores, and access policies. Teams can search, discover, govern, and collaborate around data from one place. The platform also uses AI to auto-generate metadata and surface context for AI agents via its MCP Server.
How does Atlan compare to Alation, Collibra, and Microsoft Purview?
Atlan differentiates through its collaboration-first, metadata lakehouse architecture and active metadata approach — pulling context automatically rather than requiring manual tagging. It scored higher than Alation and Collibra in the 2024 and 2025 Forrester Wave for Enterprise Data Catalogs, and has a 75% competitive win rate. Gartner named it a Leader in both Metadata Management Solutions (2025) and Data & Analytics Governance (2026).
What is Atlan's MCP Server?
Atlan's MCP (Model Context Protocol) Server pipes the enterprise context layer — all data definitions, lineage, quality scores, and business glossary terms — directly into AI agents and LLMs. This means any AI tool connected to Atlan's MCP Server instantly knows your organisation's data dictionary, business logic, and trusted data sources, enabling reliable AI reasoning in production without hallucination.