Reimagining celebration commerce: modernising the platform and changing how product and engineering work, with AI built into the process.
Current role
Chief Product & Technology Officer, FNPPreviously
Shortening the distance between a business problem and a working answer.
For over a decade, I’ve built technology across consumer businesses, from the search, discovery and checkout experiences people use every day to the supply chains and operations that make them work.
Across all of it, the biggest shift has been speed. AI now lets a small team take an idea from whiteboard to working software in days, not quarters. Knowing which idea is worth that effort is still the hard part, and it’s the part I enjoy most.
Perspective
Trying an idea now costs very little. Choosing the wrong one still costs a lot.
With today’s AI tools, a small, experienced team can go from a question to working software in days. That changes how a technology organisation should be shaped, what it should build itself, and how quickly the business can learn. It doesn’t remove the need for judgement about what to build.
Task
Until recently
With AI in the loop
First working version of an idea
One or two sprints
A day or two
Getting to grips with an unfamiliar system
Weeks of onboarding
Hours of guided exploration
Weighing three approaches
Debate, then pick one
Build all three, then compare
Internal tools for operations teams
Waiting behind the roadmap
Built alongside it
Product decisions
Prototype first, then write the spec.
Putting a working version in front of customers or operators settles questions that documents can’t. The spec gets written once we know what we’re building.
Engineering
Agents in the daily workflow.
Understanding legacy code, migrations, test coverage and reviews, with clear standards for what still needs close human scrutiny.
Operations
Automating the long tail.
Most businesses carry manual work that sits between systems. AI now makes it worth automating: catalogue work, support, partner workflows and internal tooling.
What hasn’t changed is the job itself: to look at the business through a technology lens, and turn the people, systems and now AI at hand into outcomes, whether that’s growth, efficiency or a better experience for customers.
Background
Building where software meets the physical world.
Marketplaces, logistics, mobility and commerce. In these products, software has to work for restaurants, drivers, delivery partners and florists as well as for the customer on the screen.
Experience
Infrastructure cost, latency, data consistency and tracking for an all-electric ride-hailing platform.
Launched District. Built the team, brought in the Paytm Insider and movies technology, and designed the platform end to end: discovery, orders, membership, payments and payouts.
Search and discovery, logistics and order fulfilment at national scale. Later, product, technology and business strategy for Xtreme.
Rivigo, Luxum (co-founder), Routofy (founding engineer), Flipkart and Strand Life Sciences. I learned the craft across logistics, e-commerce and scientific software, and learned what it takes to start something from nothing.
Education
B.Tech & M.Tech in Computer Science.
How I work
A few things I hold to.
Start from the outcome.
Technology earns its place when it moves something that matters to a customer or to the business.
Small teams, real ownership.
Context and autonomy rather than process. With AI, a small, senior team can do far more than it once could.
Prototype, then decide.
Evidence from a working version is more persuasive than arguments over a document.
People stay accountable.
AI drafts the work. People review it, test it and own what ships.
Notes
Things I’ve written while building.
Short pieces on engineering, product and the business side of technology, first shared on LinkedIn.
AI · Business models
AI creates value when it strengthens the core business.
The advantage isn’t the model on its own. It comes from how AI makes recommendations, efficiency and engagement better over time.
Read on LinkedIn ↗Transformation
Modernisation should make the business faster.
Faster launches, dynamic pricing, personalised discovery and smarter delivery are what a platform rebuild is ultimately for.
Read on LinkedIn ↗Product · Search
Search should understand intent, not just text.
Entity suggestions, phonetic correction, low latency and early availability signals all help customers reach what they wanted with fewer dead ends.
Read on LinkedIn ↗Marketplaces
Look at ranking through a business lens.
Visibility affects sellers, customers, ads, revenue and cost all at once. Optimising one metric on its own rarely helps.
Read on LinkedIn ↗Engineering
A bad schema can cost you 10×.
Search quality and performance are decided long before a query runs, in the choices about fields and analyzers.
Read on LinkedIn ↗Leadership
People remember how it felt to work with you.
Credit in public, hard feedback in private, and understanding the context before judging someone.
Read on LinkedIn ↗Let’s talk
Usually the best ideas come out of a conversation.
You don’t need an agenda. Thirty minutes is usually enough to find out whether there’s something worth exploring together.
Especially if you’re
- A founder or CEO working out where AI really changes your speed or cost structure
- A technology leader rethinking how your team should be shaped for the next few years
- Building in commerce, marketplaces, logistics or mobility, and want an outside view