Nandni Srivastava.

Same signal,
different answers.

Hi! I'm Nandni. I take ambiguous, messy inputs - funnels, tickets, experiments, unlabelled behaviour and turn them into whatever a team actually needs next: a metric, a decision, a shipped test or a peaceful Saturday :)

PRODUCT ANALYTICS · B2B SaaS · CONSUMER AI
raw signal clean model decide a dashboard a call to make a recommendation
01 / about

The long version

Can analyzing industry trends really drive a 55% increase in revenue? Apparently, yes!
That was one of my first real analytics wins at Quora, and probably when I realised I really liked the detective work behind the numbers.

Since then, I've worked across fintech, gaming, and quick commerce, with a slightly unusual detour into UI/UX work with an early-stage startup, pre-launch, founded by a Stanford GSB Botha Chan Fellow.

I like getting close to the actual problem; digging through behavioural data, designing experiments, building the metrics underneath them, and occasionally figuring out why two perfectly reasonable numbers disagree (Twyman's Law).

These days, I'm going deeper into statistics & experimentation. Outside work, I sing, play piano, and occasionally teach.

I write teardowns and notes as I go, mostly to think in public.

$6.2M
inventory recovered via a 0→1 returns workflow
50 → 18,000
on-demand workers scaled on a real-time payout engine
92%
faster invoice lookups after rebuilding the pipeline
Currently
Senior Data Analyst @ Blinkit
Focus
Product Analytics · Experimentation · Causal Inference · Statistics
Stack
SQL · Python · dbt · Airflow · Amplitude · Tableau
02 / experience

Where I've done the work

2024 - now

Senior Data Analyst · Blinkit

Warehouse manpower
  • Replaced HR's manual payout math with a real-time, API-driven payout engine - the transparency to scale from ~50 to 18,000 on-demand workers pan-India.
  • Shipped a predictive manpower model (order forecasts × absenteeism) on Airflow that replaced manual scheduling and cut slot-creation latency 93% (2 hrs to 8 min).
  • Built a Return-to-Vendor workflow from 0 to 1: a historical-inward ETL that dropped invoice lookups under 3s and execution from an hour to 5 min (-92%), unlocking $480K in returns in the pilot's first quarter.
  • Scaled RTV from 1 to 100 cities with cross-city billing logic - automating 1.2M+ units and recovering $6.2M in six months by extending it to expired and damaged stock.
2023 - 24

Data Analyst, CX · Jupiter Money

  • Went from shadowing a senior analyst to owning CX, Home Page & Rewards in 7 months.
  • Built 10+ dbt models for CX & Credit Cards, centralizing the logic dashboards kept re-deriving.
  • Ran A/B tests and deep-dives for high-balance users that lifted CSAT +5% and cut tickets 16%; automated a Home Page mailer that improved CTR ~7% across 15+ offerings.
2023

Data Analyst Intern · Mobile Premier League

  • Analyzed 45M+ Rummy users in SQL & Python, tracking DAU, retention, churn and revenue through Tableau funnels to surface growth levers.
  • Automated app-version adoption tracking across Databricks, Slack and Sheets, replacing a manual report with real-time visibility.
2021 - 22

Data Analyst Intern · Quora

  • Analyzed ad-industry trends in SQL & Python to find revenue gaps, contributing to a 55% lift in B2B ad revenue ($32K).
  • Built a Tableau sales tracker with APAC teams that saved 8 hrs/week of manual reporting.
03 / project & writings

Notes & things I've built

04 / contact

Have a messy question?
Let's talk.

Email Linkedin ↗ GitHub ↗ Resume ↗