I Spent 6 Years Building. 2025 Was the Year It All Paid Off
Looking back at 2025, it wasn't the smoothest ride, but it was the year a lot of things we'd been building for years finally started paying off.
A few moments that stand out:
The Internal Tool That Wasn't
What we initially developed as an in-house solution to our own issues became a desired thing in other brands. We have launched it to a few people, and the response has been in very humbling way. As it happens, when you work out your pain intensely enough, it will pierce others as well.
AI Becoming the Default
AI ceased to be an experimental tool, and it became the rule of thumb. Velocity went up. Burnout went down. Juniors began delivering work that would previously have had to be accompanied by seniors. The shift did not make a noise; it just occurred, week after week.
Watching an Industry Evolve
This year, the Indian e-commerce has been quicker than I have ever witnessed. Trading by talking, paying by agents, video shopping--all of that was no longer only a possibility, but a promise. You must have been blinkered over the transition.
The Hard Parts
The toughest moments? Two hard choices that we put off about people. Such discussions were painful, but made the way for all the rest to prosper. At other times, the right call can be a horrible one at the time and the obvious one afterward.
The Best Parts
It was so great to see the team rise in a manner that I would never have imagined. Individuals developing out of the positions they had never imagined they could hold. It is what makes the grind worth it.
2025 taught me something important:
Real progress rarely feels dramatic in the moment. It doesn't show up as a viral launch or a funding announcement. It feels like small, consistent bets that compound when no one's watching.
It's the infrastructure you build when no one cares yet.
The systems you refine when they're still clunky.
The people you invest in before they become irreplaceable.
None of it makes for a good tweet. But all of it matters when the moment finally arrives.
What I'm taking into 2026:
- Trust the compound effect. The boring, repetitive work of today can become a breakthrough tomorrow.
- Make the hard people decisions faster. Culture debt is still debt.
- When something works internally, don't be afraid to share it externally.
- Progress isn't about feeling productive every day. It's about looking back after a year and realizing you're somewhere completely different.
Grateful for every person who trusted us with their challenges this year—clients, teammates, and founders who shared their struggles in DMs. Those conversations shaped more than you know.
Here's to making 2026 even more intentional.
What was your biggest lesson or win from 2025?
Really appreciate this reflective post, Saurabh. The way you describe progress as a series of small, often unglamorous decisions that compound over years definitely resonates. It's similar to our experience: an "internal" tool we built to solve our own pain ended up being useful to others, and we only realised it by sharing it outside the company.
I also liked your point about AI becoming a default rather than a headline. The most meaningful shifts are often the quiet ones that free up people to do higher‑level work.
Would love to hear more about the internal tool you mentioned and how you decided to open it up—was there a particular moment or customer feedback that tipped the scales?
Thanks, this resonates.
There wasn’t a single tipping point for the internal tool. It emerged from repeatedly fixing the same coordination issues for ourselves. Over time it just became invisible and essential to how work moved.
When something fades into the background and removes friction without needing explanation, that’s usually the real signal.
Still early, but paying close attention to where this leads.
The "internal tool that wasn't" pattern resonates strongly. We've seen the same thing - when you build something to scratch your own itch intensely enough, you end up mapping terrain others didn't know existed. The best products often start as infrastructure someone built because they couldn't not build it.
Your point about AI becoming "the default" without fanfare is something I've been thinking about too. The loudest AI conversations are about replacement and disruption. But the actual shift I've seen is quieter: it's not "AI vs humans" but "humans with AI vs humans without it" - and the gap is widening in ways that don't make headlines.
The hardest part to internalize: "Real progress rarely feels dramatic in the moment." We're trained to look for milestone moments, but the actual work is just... Tuesday. And then Wednesday. And then you look back after enough Tuesdays and something has compounded.
My biggest 2025 lesson was adjacent to yours: the difference between "explaining a feature" and "watching it get used." If you're spending more time selling internally than observing users succeed, something's wrong with the foundation. That signal usually shows up in conversations before it shows up in metrics.
What was the internal tool that became external? Curious what the pain point was that turned out to be universal.
This really resonated. Especially the line about “Tuesday, then Wednesday”, that’s exactly how it felt.
I agree with you on AI too. The loud conversation is about replacement, but the real shift is subtler. It’s about raising the baseline. People doing better work with less friction, and the gap widening quietly.
Your point on “explaining a feature vs watching it get used” hits home. We saw the same signal internally. The moment we spent more time explaining why something existed than watching it disappear into daily workflows, we knew we were on the right track.
On the internal tool: it started as a way to reduce invisible coordination overhead across teams and stores, the kind that doesn’t show up in dashboards but shows up in delays, rework, and missed handoffs. Once it removed that friction for us, we realised the pain wasn’t unique.
Not ready to package it as a product yet, but conversations like this are helping us understand what’s universal and what’s contextual.
Appreciate you sharing such a grounded perspective.
Coordination overhead is one of those pains that's almost invisible in the moment but compounds brutally over time. The fact that you identified it as "doesn't show up in dashboards but shows up in delays, rework, and missed handoffs" tells me you've lived it deeply enough to see the second-order effects.
The "universal vs contextual" question is the right one to sit with before packaging it. The founders who rush past that question end up building features their first 10 customers needed but their next 100 don't. The discipline to have conversations like this first—to let the pattern emerge across different contexts—is rare.
One signal I've found useful: when prospects describe the problem back to you using different words but pointing at the same underlying friction, you're probably onto something universal. When they nod but then ask "can it do X?"—where X is specific to their context—you might still be too early.
Sounds like you're in the right posture: building context, not rushing to market. That patience usually pays off in product-market fit that doesn't require constant explanation.