There’s a particular kind of corporate relationship with startups that has always struck me as mildly absurd. The large company watches the startup from a safe distance, sends representatives to innovation conferences, and periodically issues a press release about being “deeply committed to the startup ecosystem.” Meanwhile, actual collaboration — the messy, operational, value-creating kind — never quite happens.
We decided to build something different.
Building the Company’s First In-House Accelerator
We setup our company’s in-house startup accelerator — and from the beginning, it was a team endeavour. Not a scouting programme, not a sponsorship arrangement, but a genuine engagement model where carefully selected startups came inside our ecosystem, worked alongside our people, and built things that connected their innovation to the customer problems we were already trying to solve.
The selection work was deliberate and collaborative. We weren’t hunting for the most impressive technology demos. We were looking for startups that aligned with specific customer needs we’d already identified together, that complemented what we were building rather than duplicating it, and that could benefit meaningfully from our scale without being absorbed by it. That last condition is harder to preserve than it sounds — and getting it right required the whole team to hold the line on it.
What the Engagement Actually Looked Like
Once selected, startups gained access to our customer network, our technical teams, and the domain expertise we’d built over years. Our engineers, product managers, and business teams engaged directly — not as reviewers or gatekeepers, but as genuine collaborators working on shared problems.
What our teams gained in return is something that’s genuinely difficult to manufacture inside a large organisation: exposure to how product and business development works when there’s no safety net. When the feedback loop is weeks not quarters. When the default answer to “should we try this?” is yes rather than let’s form a working group. That cross-pollination worked in both directions — the startups got better at enterprise-scale thinking, and our teams got better at moving with intent and tolerating productive ambiguity. Both are capabilities that compound quietly over time.
The Outcomes
Some of the startups we worked with have since become unicorns or are well on their way. We won’t claim the credit for that — great startups become great because of their own founders and teams. But being part of an ecosystem that contributed to their trajectory, and watching our own people grow through the experience, is something we consider a genuine outcome of building this the right way.
Among the partners we engaged closely were Yellow.ai and Moglix — both of which went on to significant scale. The through-line in those relationships was consistent: we aligned on specific customer problems first, and the innovation followed from that alignment rather than the other way around.
What We’d Tell Anyone Building One
If there’s one thing we learned from running this, it’s that the quality of a startup accelerator is determined less by the startups you select and more by the seriousness with which your own organisation shows up to the engagement. Startups can tell, very quickly, whether a corporate partner is there to learn and build or there to observe and claim credit.
We showed up to build. And that, more than any specific programme design, is what made it work.
What would change in your organisation if startups were treated as genuine collaborators rather than vendors or spectators?
Let’s keep learning — together.
harish.balasubramanian 
Share your thoughts