AI PRODUCT STRATEGY

Know which AI idea is worth building before you build it.

I help owners and teams decide where AI actually fits, then test the idea with a working prototype and real users. You get a clear answer on whether to build, change direction, or stop, before a full build budget is spent.

Book a Call Help Me Choose

Also open to full-time Product Owner and Strategy roles.

IdeaThe problem worth fixing
SpecScope and acceptance criteria
PrototypeA working version, real users
DecisionBuild, change, or stop

Not sure where to start?

Answer two questions and I'll point you to the engagement that fits.

1 / 2Where are you right now?
2 / 2What would help most right now?

Services

Three ways to work together, all led by strategy. Each starts with a conversation and a scope agreed before any work begins.

How an Idea Validation engagement runs

01

Define

Pin down the problem, the user, and the smallest version worth testing. You get a written spec with scope, acceptance criteria, and what is deliberately left out.

02

Prototype and test

I build a small working prototype and put it in front of a few real users. It is built to answer the question, not to run your business. Data handling follows India's DPDP Act 2023 from the start.

03

Decide

You get what the pilot showed: what worked, what did not, and a clear recommendation. If the idea holds up, the full build is done by your own engineers or a development partner.

Commercials. Fixed fee for each engagement, scoped before work starts. Ongoing support can be added as a monthly retainer.

Work

Products I have built myself, and earlier products I led.

FitCheck Live

An AI job-fit analyzer for senior professionals. It scores profile alignment, flags skill gaps, and suggests improvements. Built to DPDP 2023 patterns, with a consent step and no data retained on the server.

Open FitCheck →
Tavlo Pilot

Invoice tracking and payment reminders for small businesses. Reminders go out through WhatsApp click-to-chat from the owner's own number. Hosted on Render with a Neon Postgres database. Currently being tested with a small group of users.

Open Tavlo →
Four affective computing MVPs Earlier work

Led product for four MVPs built on FACS (Facial Action Coding System) based affective computing, which reads facial expressions to measure emotional response. I turned an experimental technology into products that could be tested with users.

DharaPutra

A bilingual crop disease diagnosis tool for Indian farmers, built on Cloudflare Workers, Workers AI, and KV.

About

The person you would be working with.

I am a Product Owner and Strategist with more than 18 years in regulated platforms, mostly insurance technology and US retirement, plus product work in affective computing and precision agriculture. I have been building and testing AI MVPs since well before the current wave of language models.

In 2023 I spent eight months studying AI systems full time, and since then I have built and shipped my own AI products. I write the spec and build working prototypes to test it, so what I recommend has been tested, not guessed. I am a product strategist, not a development agency: when an idea holds up, building it for real is work for your engineering team.

I build prototypes by directing AI coding tools, not by hand-writing every line. My part is the product decisions and the review: a multi-model pipeline drafts the spec, builds, and reviews the code, and automated tests and security scans (pytest, ruff, bandit) check the result before anything reaches a user.

  • Experience18+ years in product
  • LeadershipBuilt and led a 10-person product team
  • CertificationsA-CSPO, CSPO
  • Based inIndia, working remotely

What I have been working on lately

Skills built over the last few weeks, each tied to something I shipped.

AI-assisted build pipelineA repeatable path from brief to PRD, build, independent code review, automated tests, and a fix loop, with a different model in each role.
Deployment on free tiersTaking a prototype live on Render with a Neon Postgres database, with invoice files stored in Backblaze B2.
Scoping an MVP under real constraintsChoosing WhatsApp click-to-chat over the WhatsApp Business API, so Tavlo could be tested with real users without a registered business.
Reviewing a live pilotFinding the gaps in a deployed pilot's user journey and prioritising fixes by outcome and evidence, not by gut feel.

Common questions

Short answers before you book a call.

What is idea validation?

Testing your idea with a deliberately small working prototype and a few real users, before anyone commits to a full build. Its job is to answer one question: should this be built, and in what form?

Do you build the full product?

No. I validate ideas and I do not develop production software. If the validation shows the idea is worth pursuing, you take the spec and findings to your own engineers or a development partner, and I can stay on as product owner if you want.

What if AI is not the right answer?

Then I will tell you. A review that ends with "do not build this" or "do this without AI" is still a useful result, and it costs far less than finding out after a full build.

How is my data handled?

Data handling follows India's DPDP Act 2023 from the start: user consent, only the data that is needed, and clear retention rules. We agree the specifics before any real user data is used.

How much does it cost?

A fixed fee per engagement, agreed before work starts. Ongoing support can be added as a monthly retainer. Book a call and I will scope it for your situation.

Do you work with businesses outside India?

Yes. I am based in India and work remotely with teams here and abroad.

Have an idea to test?

Book a call, or send me a few lines about what you want to automate. I am also open to full-time Product Owner and Strategy roles.

Book a Call Describe Your Idea Connect on LinkedIn