Atlan emerged seemingly out of nowhere to develop into one of many preeminent suppliers of information catalog options. However the path to success for Atlan didn’t arrive spontaneously, and was the results of laborious work and expertise of CEO and co-founder Prukalpa Sankar, who can also be a BigDATAwire Individual to Look ahead to 2025.
BigDATAwire: First, congratulations in your choice as a 2025 BigDATAwire Individual to Watch! Again in 2012, you and your eventual Atlan co-founder, Varun Banka, have been constructing an enormous knowledge platform for prime minister of India. Did you ever suppose that work you have been doing at SocialCops would result in a profitable firm?
Prukalpa Sankar: Completely not – and but, wanting again, it feels virtually inevitable. On the time, we weren’t optimizing for achievement. We have been optimizing for impression. We didn’t got down to construct an organization – we got down to clear up significant, high-stakes issues.
From counting buildings with satellite tv for pc imagery to converging 600+ messy knowledge sources, SocialCops gave us a front-row seat to a number of the most painful, chaotic, and guide knowledge challenges on this planet. And if you dwell by that ache lengthy sufficient, you both give up – otherwise you construct one thing higher. Atlan was born out of that “sufficient is sufficient” second.
We weren’t attempting to construct a startup. We have been simply obsessive about fixing the issue the proper approach.
BDW: Atlan has develop into one of many high knowledge catalog suppliers over the previous few years, and was the far and away chief in the latest Forrester Wave for Enterprise Information Catalogs. What do you attribute that success to?
PS: Our largest aggressive benefit is care.
At Atlan, we function with a core precept: prospects > firm > group > me. That hierarchy shapes each resolution, each line of code, each roadmap debate. We really care – about fixing actual issues, about making our prospects heroes of their organizations, about being an actual accomplice of their journey.
This stage of empathy has helped us construct belief. It’s why we’ve constantly been the top-rated answer throughout industries and buyer assessment platforms. It’s additionally why we’ve been in a position to innovate forward of the curve.
We have been the primary to launch Atlan AI. The primary to operationalize Information Mesh and Information Merchandise in a catalog. We pioneered Lively Metadata and redefined the class – not as a documentation instrument, however as a residing, respiratory cloth of the fashionable knowledge stack.
We didn’t simply discuss “shifting left.” We constructed workflows that combine metadata natively inside engineering instruments. Each a kind of bets got here from listening deeply and caring intensely.
And that care shall be our edge going ahead. As our prospects face the largest shift of their careers on this new AI-native world, they gained’t want simply one other vendor. They’ll want a accomplice they’ll belief. We plan to indicate up with the identical stage of care, empathy, and innovation they’ve all the time recognized us for.
BDW: Information governance is difficult. What’s the one most essential factor that practitioners do to enhance their odds of success, or a minimum of decrease the ache?
PS: Begin with the enterprise downside. Not the know-how.
After working with 200+ knowledge groups, we’ve constructed one thing we name the Atlan Approach – a set of hard-won classes about what really makes governance succeed. Not simply the tech, however the individuals, this system, and the working mannequin.
Most governance packages fail for one among three causes:
- They by no means stand up and working.
The metadata stays dry. Implementation is just too guide. It’s too laborious to keep up. That’s why we constructed Atlan to be automation-first and to shift left – deeply integrating into the info producer workflow. Governance shouldn’t be a one-time setup. It ought to be a sustainable, long-term behavior – a part of the way you construct and ship knowledge merchandise on daily basis. - They by no means get adopted.
That is the place our change administration philosophy kicks in: don’t pressure it. Take know-how to your customers – don’t carry your customers to the know-how. That’s why Atlan exhibits up the place your group already works: inside Slack, Microsoft Groups, BI instruments, and knowledge warehouses. We meet individuals the place they’re, not the place we want they’d be. - They’re not future-ready.
Change is the one fixed within the knowledge ecosystem. Two years in the past, no person was speaking about vector databases. Final yr, they have been all over the place. This yr, the dialog has already moved on. Governance programs can’t be brittle. That’s why we’re constructing a totally open platform – so governance doesn’t gradual groups down, it units them free.
On the finish of the day, we consider governance ought to be invisible. It shouldn’t really feel like management. It ought to really feel like enablement. Embedded within the workflow. Constructed for actual people. And all the time evolving.
BDW: Atlan’s technique is to function the metadata management airplane, sitting above the info instrument stack to control knowledge through metadata. That’s not how knowledge practitioners are accustomed to doing all the things inside their instrument. What’s the secret to altering these previous habits?
PS: The key is straightforward: you don’t change conduct—you design round it.
One in all our earliest classes at SocialCops was that folks revert to what’s best. You’ll be able to’t brute-force new workflows. So as a substitute of attempting to combat that, we constructed Atlan to be the connective tissue – not a brand new silo. Our philosophy is to meet individuals the place they’re, not the place we want they have been.
That’s the place Lively Metadata is available in. Most metadata platforms act like passive libraries – nice for documentation, however disconnected from actual work. We flipped that mode. Atlan prompts metadata throughout the stack – embedding it into instruments groups already use: GitHub, Slack, Groups, dbt, BI instruments, and knowledge warehouses.
We’ve introduced metadata into engineering workflows, the place producers really construct and ship knowledge merchandise. We’ve helped knowledge customers discover trusted context proper contained in the instruments they already use. That is what we imply by shifting governance left – governance that looks like a characteristic, not a friction.
As a result of on the finish of the day, “Metadata isn’t a layer you add. It’s the muse you construct on.”
BDW: GenAI instruments and LLMs are proliferating in enterprise knowledge stacks. What difficulties do these new instruments and applied sciences pose to knowledge governance?
PS: We’re not in a digital-native world. We’re coming into an AI-native one.
Essentially the most fascinating factor about LLMs is that they now perceive language – however they don’t perceive which means. Solely people can train that. And as LLMs begin doing extra of the work people as soon as did, one query issues most: are you able to belief it?
Are you able to belief the info that skilled the mannequin? Are you able to belief the mannequin that produced the output? Are you able to belief the AI-generated motion that impacts what you are promoting, your prospects, or your model?
That’s the place governance steps in. Not as coverage enforcement, however as a system for context and belief.
Within the AI-native enterprise, governance isn’t a back-office perform. It’s a frontline enabler. The businesses that transfer quick and construct belief would be the ones that win. However that’s solely attainable if governance evolves into an clever, embedded, real-time functionality.
We consider that is governance’s leapfrog second – an opportunity to maneuver from being a value heart to a aggressive benefit. As companies rewire their merchandise and processes with GenAI, the actual query gained’t be “Can we do that?” Will probably be “Can we belief this?”
That belief must be systemic. It will possibly’t cease on the knowledge. It has to stream by your complete lifecycle of selections, fashions, and automation. That’s the function of Lively Metadata as a semantic layer: making which means machine-readable, making governance invisible, and serving to AI act with context and care.
And that’s why “Within the AI-native period, governance isn’t a blocker. It’s the unlock.”
BDW: What are you able to inform us about your self exterior of the skilled sphere – distinctive hobbies, favourite locations, and so forth.? Is there something about you that your colleagues is perhaps shocked to be taught?
PS: I’m the one Prukalpa on this planet – actually. My dad and mom say they considered web optimization earlier than Google existed, and actually… they weren’t improper.
To learn the opposite BigDATAwire Individual to Watch interviews, click on right here.
