Sumeet Gadodia.

Hi, I’m
Sumeet.

Product person. Builder. Curious optimist.

I enjoy turning complex problems into simple products — from commerce and AI at Aza to things I’m building myself.

Product
Thinking
AI Product
Builder
Mentor &
Team Builder
Mumbai &
Beyond
Let’s connect Explore my work
Sumeet at a desk with a Mumbai skyline behind him
PRODUCT DISCOVERY✳AI IN THE REAL WORLD✳POST-ORDER EXPERIENCES✳BUILDING FROM ZERO✳PRODUCT DISCOVERY
01 / Currently building

My current builds.

Two ideas I’m turning into products. The interesting part is what I’m learning along the way.

Nirmit.life

AI video creation for India.

Turn a photo and your words into lifelike videos — in your voice, your language and your style.

  • Talking avatars
  • Multiple Indian languages
  • Simple, affordable creation
  • For creators, educators and SMEs

On the workbench · Sep 29 Making a first video feel easy: person, voice, script, then video. I’m also comparing avatar quality with the cost of creating it.

Have a video use case? Let’s talk ↗

Conceptual visual · Product in development

JoshFast

AI-first customer support for modern teams.

What would customer support look like if AI wasn’t bolted on to an old ticketing system, but designed into the product from day one?

  • Conversations across channels
  • Knowledge and playbooks
  • AI routing and approvals
  • Human in the loop

On the workbench · Sep 20 Giving support decisions an owner through playbooks, routing and human approvals.

Building support workflows? Let’s talk ↗

Conceptual visual · Product in development

02 / Notes from the workbench

My thinking, in the open.

Questions, experiments and lessons from products I’m building and problems I’ve already shipped.

01 / NIRMIT

Which video model survives a real use case?

Use the same input to compare identity, lip sync, language, latency and cost. Choose for repeatable quality and viable economics.

02 / SUPPORT AI

When should an agent ask a person?

Automation works best when escalation is designed, not improvised. Define confidence, exception and risk thresholds so the model knows when to stop.

03 / CATALOG AI

What does good enough mean?

AI quality is not about eliminating every error. Evaluate the mistakes that affect customer trust, discoverability or downstream operations first.

04 / POST-ORDER

Which system changed the date?

A delivery promise can fail far from the UI. Trace each system that writes or overrides the date before fixing what appears to be a front-end problem.

05 / MARKET TEST

Can you test before scaling?

Test the assumption with a small, controlled experiment. Use one-time translation to learn whether the experience is useful before investing in scale.

Mumbai street scene at golden hour
Build for the people in the context they're actually in.
03 / Product work

Problems I've spent time on.

Commerce, service operations and the decisions behind a better customer journey.

13+ yearsFrom 0→1 products to scaled platforms
~1.5MMonthly visits to Aza’s marketplace
10×Onsitego service requests, 10K→100K/month
40 → 70Onsitego Net Promoter Score (NPS)
01 / CATALOG AI

Better product data at a fraction of the cost.

Built an AI-powered enrichment pipeline to standardize, classify and enrich catalog data — reducing cost from ₹22 to ₹3 per product with ~3.5× productivity.

Extract attributesEvaluate qualityReview exceptions
Illustrative fashion commerce team reviewing catalog and service workflows
AI in Commerce · Video series

Behind the build.

The stories above also shaped how I think about AI. Here are three short videos on what worked, what failed and what changed.

EPISODE 01 · VIDEO

What work should AI take off people’s hands?

Moving beyond chatbots to useful work across support, catalog and sales.

LinkedIn ↗
EPISODE 02 · VIDEO

The prompt was only part of the problem.

Why context, rules and workflows matter when AI joins a business process.

LinkedIn ↗
EPISODE 03 · VIDEO

When AI is confidently wrong.

Checks, feedback and human approval around product classification.

LinkedIn ↗
AI in Commerce
An evening in MumbaiMumbai, where I call home
04 / The path here

What’s the next harder problem worth solving?

That question has taken me across enterprise software, commerce and the systems that keep a customer promise. 13+ years in, I’m still following it.

Mumbai is home, with Pune for college and brief work stints in Delhi/Noida and Bengaluru.

Zepo

Where I found product management.

Building ecommerce tools for sellers led me into logistics too, with Zepo Couriers. The operational problem was often bigger than the software screen.

Onsitego

Learning to build behind the scenes.

CRM, support, partner workflows and field operations taught me how much of a good customer experience depends on systems people never see.

Aza Fashions

Rebuilding, then scaling.

From the ecommerce platform and discovery journey to post-order experience and profitable growth, the work kept crossing team boundaries. I built order, cancellation, return and refund flows from 0→1.

Now

Making AI useful in everyday work.

At Aza, I lead work on support copilots, sales agents, catalog intelligence, visual merchandising, taxonomy and voice of customer — bringing AI into operational decisions with room for human judgment.

I’m still curious about the next hard problem. If something here sparked a thought, I’d enjoy hearing it.

05 / LET’S CONNECT

Tell me what you’re trying to solve.

Whether you’re hiring for a product role, building something new, or simply comparing notes — I’m always open to a useful conversation.

Looking for a product leader?

I work across AI-enabled products, commerce, Search & Discovery, Growth, CX and platforms — from 0→1 to scale.

LinkedIn ↗WhatsApp ↗Email ↗