Eight years
in analytics.
Bengaluru. Walmart's shoppers, modelled from here. Flipkart's delivery promise. A lender that became a bank, and every ledger in it made to agree.
Previously: a 10-person pod, four business pivots, and a bank's worth of ledgers.
登場人物tōjō jinbutsu Characters Characters
Bhavuk Manocha
The bottleneck isn't data anymore. It's the loop between asking and acting.
Profile
- Base
- Bengaluru
- Studied
- Chemical engineering, IIT (ISM) Dhanbad
- In analytics
- Eight years
- Last post
- Analytics Manager, Slice
- Now
- Co-founding a company in stealth
- Off the clock
- Badminton, Strategy and RPGs, Reading, Travel
得意技tokui-waza Signature moves Signature moves
- Data systems
- AWS (S3, EMR, EC2), PySpark, Airflow DAGs, Delta Lake, ETL pipelines, Warehouse tuning
- Analytics
- SQL, Python and pandas, Superset, Tableau, A/B testing, Financial modelling
- Fintech
- Lending analytics, RBI regulatory reporting, Financial reconciliation, Unit economics (LTV, CAC, MAU)
- Leading teams
- Hiring across an analytics org, Mentoring and promotions, Budget ownership, Crisis management
- AI tooling
- LLM integration, Cursor with custom rules, Retrieval (RAG), n8n automation
既刊kikan Volumes in print The series
Whatever goes in has to come out somewhere.
IIT (ISM) Dhanbad, the prequel
B.Tech, Chemical Engineering. Mostly about making a system's inputs and outputs agree, which turned out to be the rest of the story.
- Four years in Dhanbad; graduated May 2018
$3,000 potential revenue per household.
Tredence, on the Walmart account
Business Analyst, then Senior, on Walmart US. Shopper analytics for Walmart US; ran the client relationship solo.
- Retail analytics for Walmart US, from the Bengaluru office
- Ran the US client relationship solo, including executive presentations
- Promoted from Business Analyst to Senior Business Analyst
One-day delivery in metros: 30% to 80%.
Flipkart, planning
Planning Analyst. Capacity planning against the one-day delivery promise in metro cities.
- Planned against the one-day delivery promise in metro cities
- Automated daily and monthly planning, saving two days a cycle
Two ledgers, one number: 0.01 tolerance.
Slice, core banking
Senior Analyst to Analytics Manager, Core Banking. A 10-person pod, through credit card, lending, payments and banking. Led the core banking analytics pod through the move from fintech to small finance bank; owned the analytics hiring charter and a $200k AWS budget.
- Defined monthly transacting users and the cohort tracking behind board-level growth numbers
- Owned the hiring charter for the analytics org: 120+ interviews, and trained other interviewers
- Owned a $200k AWS analytics budget; auto-terminating EMR clusters kept the burn in check
- Promoted two people from the pod
- Named AI Pioneer twice
- Promoted from Senior Analyst to Lead Analyst to Analytics Manager
Co-founding a company in stealth. More soon.
Still in the shrink-wrap
No preview pages yet. Read the next-volume notice.
目次mokuji Table of contents Contents
Ch. 1The expecting-mothers journeyVol. 1
$3,000 potential revenue per household. Modelled how expecting mothers shop for Walmart US, worth about $3,000 of potential revenue per household.
Ch. 2One-day delivery planningVol. 2
30% to 80%. The one-day delivery SLA in metro cities went from 30% to 80%.
Ch. 3Peak-sale fixesVol. 2
7 days to 1. Automated diagnostics cut the time from spotting a problem to fixing it during peak sales from seven days to one. Daily and monthly planning was automated too, saving two days a cycle.
Ch. 4RADAR, finance alertingVol. 3, 2023–24
₹1.2L a day caught. One alerting framework for finance: 150+ alerts automated, false positives down 80%, issue detection 90% faster, and ₹1,000 Cr+ of historical accounting issues surfaced and resolved.
Ch. 5Accounting V2, two-way reconciliationVol. 3
0.01 tolerance. Loan-management and core-banking ledgers reconciled in both directions, closed T+1. SLAs up 150%, downtime down 60%, and 20+ Airflow DAGs at 99.99% on ₹50M a day of settlements across 10+ entities.
Ch. 6TDS compliance under section 194NVol. 3
₹3.04L refunded. Detection, refunds and tracking, end to end: 100% accuracy, zero penalties, about 30 hours a month saved.
Ch. 7Investor databookVol. 3
$30M debt raise. Automated the due-diligence reporting behind a $30M debt raise. Weeks of work became hours, with no manual errors.
Ch. 8AI across analyticsVol. 3, 2024
100+ hours a month. Co-founded the central AI pod and ran the first Cursor rollout with custom rules, plus Compliance GPT for checking marketing copy against RBI guidelines. 29 tickets resolved by AI, first response from an hour to 25 seconds, and the AI Pioneer award twice.
次巻予告jikan yokoku Next volume preview Next volume
おまけomake Bonus pages Extras
Weekend rallies, mostly.
Reasonable forehand.
Slow campaigns, fast saves.
Currently rotating.
Non-fiction by day, fiction by night.
Margins always written in.
Always mapping the next trip.
Open to a window seat.
作品データsakuhin dēta Series data Series data
| Title | Bhavuk Manocha, the series |
|---|---|
| Status | Ongoing |
| Setting | Bengaluru |
| Serialized since | 2014 |
| Volumes in print | Four, and one in the shrink-wrap |
| Chapters | Eight |
| Largest cast | A ten-person pod |
| Plot twists | Four business pivots |
あとがきatogaki Afterword Afterword
Thanks for reading this far. Volume four is being written as you read this, and I can't show you the pages yet. If you want to talk about what's in it, or about ledgers that refuse to agree, write to me.
読者コーナーdokusha kōnā Readers' corner Readers' corner
To Bhavuk Manocha
Bengaluru
Letters about any volume are welcome, including the one still in the shrink-wrap.
manocha.bhavuk@gmail.com
奥付okuzuke Colophon: the printing details at the back of a book Colophon
Bhavuk Manocha, the series. First edition, 2026. Colour page: the sky over Bengaluru, as it is right now.
Published by the author, Bengaluru. Typeset in Zen Kaku Gothic New, Shippori Antique B1, Dela Gothic One and Shippori Mincho.