Machining multiple missing context layers for modern AI-forward enterprises

Founded 2013

Partnered 2018

Currently Series C

Two young individuals, who had never worked a formal, full-time job, were obsessed with creating a world-class company. Their eyes twinkled when they explained how data can alleviate poverty, impact healthcare and impart education. The kind of twinkle that stays with you long after the meeting is over.

So they started SocialCops. We backed them when it was still a consulting company. By definition, it was not VC-backable. But their work was phenomenal.

They partnered with governments and federal agencies, dealing with diverse data sets, ranging from geo-mapping 500M Indians to processing billions of pixels of satellite imagery. They had touched 100M lives already, if not a billion.

What wasn’t rosy was the everyday firefighting with messy data. They had instances where certain numbers on a data dashboard would be broken just 15 minutes before a meeting with the Prime Minister’s Office, and they’d have no idea why.

They were not alone. Data is scattered everywhere, more so in enterprises. Humans are diverse, so are their working styles and associated tool stacks.

They needed a collaboration plane. That was v1.0 of contextual intelligence, prompted back in 2018.

Prukalpa Sankar

Varun Banka

Prukalpa Sankar & Varun Banka · Atlan

"WaterBridge was our earliest supporter, even before Atlan existed. I always tell Manish that I have no idea what he saw, when he invested in us back then. Without WaterBridge, we would not have been able to build everything that we have so far. From the get-go, they respected our time, moved quickly and began to add value even before they invested."

Total Raised $215M

Fortune 500 Customers 300

Customers’ Market Cap $10T+

Competitive Win Rate 75%

Why we backed it

Data teams have a diverse DNA. So, they need a collaboration plane to function

2018. Big Data was popular. Data teams held the maximum sway. But it was chaos. Coders, analysts, managers and leaders had unique tool preferences. Companies needed a collaboration plane, but none existed. To Prukalpa and Varun operating deep in the trenches, it was clear as daylight.

01 · Founders

Career entrepreneurs scarred from dealing with messy data

NTU classmates who, while on a college bus ride, had the urge to leverage data for good. Decided to become entrepreneurs. All while they were 21.

02 · Market

P = f (I,C)

Performance (P) is a multiplicative function of Intelligence (I) – commoditised – and Context (C ), the connective tissue of tacit knowledge. Intelligence is contextual, the way walking is gravitational. And context is king.

03 · Wedge

Be iconic. Build the best product in the world. Nothing less.

The intent from Day 1 was to search very hard for signals of truth, not traction. 'Never done before' is not an excuse. Defy all limitations, deny all odds.

From a scattered consulting company to a category creator

2013 - 2017

Encoding: the first impression with SocialCops

SocialCops mapped India's 640K villages to deliver LPG connections to 70M below-poverty-line women under the Ujjwala Yojana scheme, created Disha - the National Data Platform that the Hon’ble PM himself uses. Built a Palantir-like company developing internal data platforms for governments, the UN and the World Bank.
2018

Cache: the WaterBridge cheque and Atlan

We led the seed after 6 months of chasing. SocialCops starts morphing into Atlan soon after. Everyone was baffled. Fresh slate, starting from scratch, in a boring space. Investors did not understand why. Buyers could not figure out the product. But users loved it. A lot.
2020–21

Long-Term Potentiation: the category starts firing.

Insight, Peak XV, Salesforce join the cap table. Atlan’s context layer sits closest to where teams work: Slack, Teams, BI tools. Snowflake IPO attracts eyeballs. Analysts name the category "active metadata management". The metadata layer becomes the wired set of neurons that power the whole modern data stack.
2023–24

Persistent Storage: the system of record for context in the AI era.

Raises $105M Series C from GIC and Meritech. Becomes a Leader in the Gartner Magic Quadrant and Forrester Wave. Partners with Microsoft Azure and OpenAI to build the first AI agent that can create its own context. Introduced Context Lakehouse - the first context infra architecture for the AI era.
2025

Total Recall: the memory machines read from.

As enterprises wire up agents, Atlan’s context layer for AI becomes existential. Becomes 2X Leader in the Forrester Wave, Customer Favourite in G2, 2X Leader in the Gartner Magic Quadrants. Complete domination across every context category. Hosts Re:Govern, the world's first conference on context.
2026

The Memory That Writes Itself: compounding context hits a tipping point.

Atlan's context lakehouse becomes the primary consumption interface for AI agents operating with context. In just two weeks across 50+ enterprises, Context agents produced 4X the volume of documentation that was written by hand in one year. 1M+ descriptions generated with 87% human parity quality, saving 110K+ hours.

A WaterBridge note

In 2017, SocialCops was a runaway success, by every bootstrapped metric. Prukalpa and Varun kept the company afloat with prize money won from every B-plan competition they participated in. The consulting company was scaling 2-3X YoY, with SaaS-like margins of 60%. Startups could only dream of government contracts. But SocialCops was already creating the national data lake for India.

Then came the time to redream. Data teams were becoming important. Varun and Prukalpa wanted to start retooling them. They had just raised their seed. The transition idea seemed unnecessary. But to take a shot at building anything iconic, the duo had to disrupt the current.

Their experiments started with 3 productised tools used internally. No GTM playbook in sight. Two products had an early head start. One had ~200 customers, and the other had $2M in revenue.

The third one, Athena, had only 1 customer. The only metric going for it was insane user love. Think Slack and Figma-level adulation. It was 2019-end. Only 9 months of cash left in the bank. The founders decided to go all in with Athena and shut the other two down. That became Atlan v1.0.

Even then, building the context plane for data teams was a baffling idea for investors. In the pre-Snowflake IPO era, this space was boring. The first few hires had bombed. Athena had no warm customer intros. Prukalpa and Varun continued building anyway.

They shifted to the Valley to pursue building an iconic company. They saw the ‘AI chasm’ coming from a distance, and created the enterprise context layer. Today, contextual intelligence is the talk of the town.

We’re proud to have backed the team when every tiny detail in our investment memo screamed risky. Today, mainstream media has made it seem like a no-brainer.