Cannot display the PDF here? Open it in a new tab
Aspiring PM · Class of 2027 · Greater Noida
I learn product management by doing the work: finding signals, making trade-offs, coordinating delivery, and reflecting on what actually changed.
The analytical base I keep reaching for when a product decision needs evidence rather than opinion.
Core technical team for the campus competitive-programming community.
Owned budgets, sponsorships and operations for the campus esports community.
Owned the visual identity and ran a full redesign across events and campaigns.
Marketing and web-traffic analytics — cleaning the data, then turning it into the dashboard behind budget decisions.
Product lifecycle and delivery across CRM, inventory and location-mapping products.
Essays on product, data and the parts of the job that never make it into a changelog — published on Medium, and written to be read by someone who does not already work with me.
The product work behind Tabyss: the charter, jobs to be done, the North Star metric I chose and the two I ruled out, twenty nine decision records, and the six finished features cut before release.
Read on MediumMost people learn the MVP and stop. There are two more questions — can someone understand it, and would they miss it — and they decide whether anyone keeps using what you built.
Read on MediumWhat actually changes when the volume of evidence outgrows the meeting it was meant to inform, and where judgement still has to do the work.
Read on MediumA short piece from before any of this — kept here because the way you notice things does not start when you get the job title.
Read on MediumA focused selection showing how I move from problem framing to requirements, prototypes, validation, and delivery — and what each one actually taught me.
Live on the Edge Add-ons store
45+ owner conversations
15+ client projects coordinatedDiscovery ◍ analytics, and technical fluency ▚ combined — turning an ambiguous problem ◈ into a decision a team can actually act on ✦ this week.
User interviews, market investigation, problem framing, journeys, assumptions, and MVP definition.
SQL, Python, funnels, engagement trends, dashboards, and evidence-backed recommendations.
Requirements, backlogs, Jira, sprint planning, stakeholder alignment, risks, and documentation.
FastAPI, Next.js, APIs, databases, cloud architecture, Git, and functional validation.
Same rigour at every size. The only difference is how much of the problem is already known when I start.
Weeks 1–2
Before scope, before tickets. Talk to the people living the problem and find out which parts of it are actually worth solving.
For problems that are still fuzzy and need a shape before anyone commits.
Weeks 2–4
Turn what I heard into something a team can build against: scoped, sequenced, and honest about what is being left out.
For teams that have signal but no shared definition of done.
Ongoing
Keep priorities, owners, and risks visible until it ships — then read what the data says and feed it back into the next cut.
For delivery that has to survive contact with real clients and real deadlines.
Every product has a next step hiding behind an unanswered question. I am good at finding which question that is.
Have something in mind?
Let's TalkNot theory. Five things I got wrong first, then corrected — each one tied to a project where it actually cost something.
It is tempting to design the polished side of a marketplace first. The 45+ owner conversations behind BenNest changed the product more than any wireframe did — half the features I assumed were essential never made the cut.
Broken instrumentation does not show up as an error. It shows up as a confident decision made on numbers that were never real. Auditing what is measured is part of the analysis, not a chore before it.
Across 15+ client projects spanning technical, 3D rendering, and sales teams, almost nothing was blocked by a lack of effort. It was blocked by nobody knowing whose turn it was. Naming the owner fixed more than escalating did.
Debating an idea in a doc is cheap but slow. Building the thin version of it is slightly more expensive and settles the question. ProdIntel AI exists because a ranked recommendation was easier to judge than to describe.
A requirements doc that only lists what is in scope is half a document. The not-now list is what stops a team from quietly rebuilding the thing you already ruled out three sprints ago.
Because the work I keep gravitating toward — framing the problem, talking to users, deciding what not to build, keeping delivery honest — turns out to be the job. The technical side makes me a better partner to engineers; it is not the part I want to optimise for.
Yes, enough to be useful and not enough to be precious about it. FastAPI, Next.js, SQL, Python, Git. I prototype to settle arguments and to understand what I am asking engineers for — not to own the codebase.
Interviews first, and with the side of the market that usually gets skipped. For BenNest that meant 45+ property owner conversations before any requirement was written. I map assumptions, mark the risky ones, and go test those.
Narrow it until a decision is possible. That usually means separating what we know from what we are assuming, picking the one assumption that would hurt most if wrong, and finding the cheapest way to check it this week.
Jira for delivery, SQL and Python for analysis, GA-style web analytics for funnels and engagement, Figma for flows, and whatever the team already lives in. The tool matters far less than whether priorities and owners are visible in it.
By what changed for the user and the business, not by tickets closed. That means agreeing on the metric and the instrumentation before launch, so the post-launch conversation is about evidence rather than opinion.
Yes — I am class of 2027 and open to product internships, associate PM roles, and project work now. Fastest way to start is email; I will reply with context on what I have shipped and where I would fit.
Going deeper on product analytics and on AI-assisted decision systems — ProdIntel AI is where most of that thinking currently lives. Alongside that, more discovery reps, because that is the muscle that compounds.
