Shivali Sharma Resume

Who am I?

Hi, I’m Shivali.

6 yrs finance , now Product Manager & AI builder

I spent six years inside spreadsheets. Turns out, the interesting problems were outside the cells.

Now I research, design and build </> them.

Finance person who knows code.

Bengaluru / 2026 Open to PM roles at early-stage AI teams

Background

Full timeline
  1. 2026The Credix part-timeFounding Product. AI receivables product, from discovery to shipped workflows.Remote
  2. 2026Rethink Systems cohort course9-week AI product management course, Cohort 8. Shipped Nibble and Loocator, live.Bengaluru
  3. 2025 – 26BiteSpeedAR / O2C Lead. Built the AR team and the finance dashboard.Bengaluru
  4. 2023 – 25SGGG Fund Services Toronto-basedFund Accountant. NAV and reporting for institutional funds, hybrid.Chandigarh
  5. 2020 – 23Lotus Design LabsFinance Associate. Month-end close, reconciliations, reporting.Chandigarh
  1. 01 · BiteSpeed finance dashboardDSO 50 → 30 daysBuilt the AR dashboard the team adopted as its daily source of truth.
  2. 02 · Nibble, AI product44 DAU → 84 WAUShipped 0 → 1 in nine weeks: discovery, PRD, evals, 255 automated tests.
  3. 03 · Loocator, field product19 spots verified on footA washroom map for women in Bengaluru, brief to live in nine days.

The one that started everything

BiteSpeed Finance Dashboard

I ran collections at BiteSpeed and got tired of asking five people for one number. So I built the dashboard myself: collections, ageing and DSO, all in one place.

It became the AR team’s daily source of truth and was used for weekly management reviews and P&L inputs.

My part: found the problem while doing the job, built the product, drove the adoption.

DSO 50 → 30 daysthe number that got me into product

Next.js·Supabase·Adopted by the AR team

Explore the dashboard
Live product demo · mock data
finance-bsd.vercel.app
Finance overview: total received, outstanding, overdue and pending totals above recent invoices and recent payments
finance-bsd.vercel.app
AR ageing report: invoices bucketed by days overdue, with brand, due date and amount
finance-bsd.vercel.app
Reconciliation: bank statement entries matched against outstanding invoices, with matched and unmatched totals
  • 1/3Overview: received, outstanding and overdue at a glance
  • 2/3AR ageing: every open invoice by days overdue
  • 3/3Reconciliation: bank entries matched to invoices

A product built from zero

Nibble

Nibble drops students into messy, realistic workplace situations. You investigate what is going on, collect evidence, work with an AI collaborator that is wrong on purpose, make a call and defend it.

My part: discovery and research, interaction design, the product definition and PRD, and building the working product.

44 DAU → 84 WAUreal usage, week one

AI product·Human-in-the-loop evals·255 automated tests

Read the case study Open Nibble
Nibble mission brief: URBN FITS, the comments are piling up and something is off; read the Instagram comments, find the pattern, decide what the team should fix
Inside the Nibble world: the URBN FITS Instagram profile with 342 posts and 128K followers, and a View problem button
Investigating with the AI collaborator: a chat with urbnfits who says they spent four lakh rupees on ads in seven days and sales barely moved
  • 1/3The mission: a real brand, a messy situation, one question
  • 2/3The world: a familiar product you already know how to use
  • 3/3Investigating with an AI collaborator that is wrong on purpose

A public-utility product

Loocator

A map of washrooms women can actually use, starting with HSR Layout, Bengaluru. Every pin carries six facts checked in person: a women’s section, a working latch, water, light, pads nearby, free or paid. Because a toilet nearby doesn’t mean you can use it.

My part: the field research, the data model, the design and the build. Every spot verified on foot.

19 spots · 6 facts eachchecked in person, not scraped

Consumer·Maps·Field-verified data

Read the case study Open the map
loocator-rho.vercel.app/map
Loocator map of HSR Layout with 19 verified washroom pins, filterable by free, public toilet, cafe, petrol pump and paid toilet
loocator-rho.vercel.app/map
Washroom detail for Karadi Cafe: verified five days ago, women's washroom yes, working latch yes, water yes, light yes, pads nearby no, free
loocator-rho.vercel.app/map
Filter washrooms by access, essentials and place type: free, walk-in without purchase, open now, working latch, water, light, pads nearby
  • 1/3The map: 19 verified pins around HSR Layout
  • 2/3A pin: latch, water, light, women’s section, freshness
  • 3/3Filters: what matters to you right now

An AI product, built with the expert

Shravan’s Round

A 20-minute mock interview for people moving into product, built with PM mentor Shravan Tickoo. Real Indian cases: Swiggy, Zepto, Flipkart, CRED. Not US textbook ones.

Your answer decides the next question, so there is no script to memorise. Hinglish is fine: clarity is what gets scored, not accent. No login, no card, free score card at the end.

My part: the persona, the case bank and the rubric, plus the knowledge layer that lets it answer the way Shravan actually teaches.

597 of his sessionscohort classes and videos, searchable mid-answer

AI product·RAG·Built with a domain expert

Read the case study Sit a round
Live product · free, no login
shravans-round.vercel.app
Shravan's Round: rehearse the round before the round, a 20-minute AI mock with real Indian PM cases and a free score card
shravans-round.vercel.app
Twenty minutes, most of it is you talking: two setup choices, Shravan asks and you answer out loud, then a score card and one fix
shravans-round.vercel.app/feedback
The score card: overall feedback, strength and growth areas, and the three things to focus on before the next interview
  • 1/3The promise: rehearse the round before the round
  • 2/3Twenty minutes, and most of it is you talking
  • 3/3The free score card: what was strong, and the one thing to fix

Experiments / playground

Things I built to answer a question

Squoobi In progress

Everything you save and never find again: reels, screenshots, repos, notes. All in one place, each one explained back to you in plain words.

Consumer·AI summaries·Building now

Building · no live link yet

Also on the shelf: a day-pass booking flow that shortened the journey from choosing a workspace to checking out.More small builds are always in progress.

About

I kept asking:
does it have to be
this messy?

Himachal-born·Defence-kid raised·Bengaluru, now

I’m a mountain girl from Himachal, raised as a defence kid. New cities, new people and starting over were part of growing up. Maybe that’s where the curiosity came from: the habit of observing first, adapting quickly and always asking one more question.

Six years inside finance taught me how decisions move through a business. At BiteSpeed, I stopped only reporting the friction and started building around it.

Now I research messy problems, design the flow, and ship the answer.

Product Manager & AI builder · Open to PM, AI PM and AI builder roles

Showing the photo

6 yrsFinance & operations
$500K+Monthly receipts
0 → 6AR team built
160+Nibble users, last 30 days

Kind words

Don’t take my word for it.

They already wrote it down.

She brings strong ownership to her work… she identifies issues early, communicates clearly with internal teams and clients, and works towards timely resolution. She's organized, dependable, and handles pressure well.
Anish MishraManaged Shivali directly at BiteSpeedLinkedIn
She is a phenomenon in what she builds, coming from the finance side. She can start from scratch and build something crazy like Nibble; she visited colleges to find and understand the deep problem. Her design sense and her urge to discover in the discovery phase are the best.
Unnati NakraCo-founder, Rethink Systems; Shivali's mentorLinkedIn
Even though I've been her mentor and wellness coach, I find myself learning something new from Shivali every day. Whether it's a small insight or a fresh perspective, she always has something valuable to share.
Harshit BadigerBuilding AI products at Headout; Shivali's mentorLinkedIn

How I think

I don’t jump to solutions.
I follow the signal.

The tools change. The thinking doesn’t.

See the illustrated version on Play
  1. 01 / DiscoverSignal

    What’s actually happening, not what’s being reported?

    Learnt · NibbleWatch what people do.
  2. 02 / DefineProblem

    Whose problem is this, in one sentence?

    Learnt · LoocatorSolve for one person first.
  3. 03 / PrioritiseChoice

    What breaks first if we ignore it?

    Learnt · NibbleFix drop-off before features.
  4. 04 / MeasureOutcome

    Which number moves if this works?

    Learnt · BiteSpeedOne number changed my career.
  5. 05 / BuildSystem

    What’s the smallest thing that proves it?

    Learnt · OrqenHarsh feedback is a spec.
  6. 06 / ShipRelease

    Who uses this tomorrow?

    Learnt · LoocatorVerify on foot, then ship.