AKANUBHA KANWAR

I prototype fast.AI does the heavy lifting.I'm more interested in the problem, friction, churn, drop-off, habit, gap.

APM at Emversity, trained in biosciences at IIT Roorkee — an unusual route into product, but one that taught me to sit with ambiguity before reaching for a solution. I've shipped tools myself, end to end, and that's changed what I look for in a problem before I commit to it.

Based inBengaluru, India
Available forAI-native PM roles · Series A through growth-stage · Remote or Bengaluru
ACROSS THE PRODUCT LIFECYCLE

How I work.

01

I find what's worth building.

I stress-test scope with data before committing engineering time. A sales funnel RCA across 300K+ leads once turned into four concrete, actionable hypotheses — that's the bar I hold every “should we build this” question to.

Discover
02

I don't wait for engineering.

When a v1 needs to exist, I build it myself. EmStudio, PitchMate, and Socrates were all coded solo using Claude, GPT, and Cursor — no hand-off required to prove an idea, and no black-box logic either.

Build
03

I close the loop with data.

Every ship gets a metric attached before it goes out, then I go find out if it moved. That's how a 24-hour contact window and a 7-minute re-escalation threshold got discovered, not assumed.

Measure
EMVERSITY · AI PRODUCTION PLATFORM

AI does the content.
I built the check.

RolePM + builder, end to end
Year2024–25
StackCursor · OpenAI API · Next.js
Read case study →
AI ProductionBuilt soloQC Engine

A QC engine for AI content production, built and shipped by me — workflow, frontend, and QC logic, end to end. It flags spelling, factual, and image errors, each with a suggested fix for a human to accept or reject.

EMVERSITY · ROOT CAUSE ANALYSIS

More leads.
Fewer students.

RoleFull analysis, end to end
Year2025–26
StackBigQuery · SQL
Read case study →
BigQueryRoot Cause AnalysisExec Reporting

A root-cause investigation across 1.4M+ leads — leads were up 85.7% while enrollments fell 23.7%. Found the collapse point.

PERSONAL PROJECT · JOB SEARCH OS

Run your job search
like a pipeline.

RoleSole builder — product, architecture, code
Year2026
StackNext.js 14 · Supabase · GPT-4o
Read case study →
Next.jsLive productBuilt solo

A job-search tracker and outreach OS — shipped in 20 days with zero prior Next.js experience.

THE REST

More work, by type.

01
AI Product · PM-led

EVA

An AI screening & interview platform — MCQ test plus a live voice interview — that screens applicants before admission.

Voice AIAdmissions
Read case study →
02
AI Product · Solo Build

PitchMate

An AI sales-training platform that role-plays real objections so reps rehearse before the real call.

AI RoleplaySales Enablement
Read case study →
03
Platform Product · PM-led

EVA 2.0

The placement operating system that runs after EVA admits a student — an ops console for admin, a self-serve hub for students.

PlacementsTwo-sided platform
Read case study →
04
Analytics · AI data product

Socrates

Two internal AI products in one — natural-language querying over the database, and behavioural clustering on sales calls.

NL-to-SQLClustering
Read case study →
05
0→1 Strategy

VR Anatomy Sim

Scoping, BRD, cost analysis, and delivery timeline for Emverse's VR anatomy simulation — Meta headset and web access paths.

VRRollout
Read case study →
ABOUT ME

Biosciences by degree. Product by choice.

I studied Biosciences and Bioengineering at IIT Roorkee, then moved into product because I liked the version of problem-solving where you actually ship the fix. I'm now looking for AI-native PM roles at Series A through growth-stage companies — the kind where being able to build the first version myself is a feature, not a workaround.

Education
IIT Roorkee · Biosciences & Bioengineering
Current role
Associate Product Manager, Emversity
Recognition
Hult Prize 2023 · Global Semi-Finalist