Atreyi Dasgupta
0%

Atreyi
Dasgupta

Physics brain. Product instincts. I find the story in the data, then figure out what to do about it.

Product Management
Data Analytics
Product Strategy
AI Agents
SQL · Python · Tableau
BITS Pilani · IIT Bombay
See My Work Let's Talk →
Scroll
Analytics✦ Product Thinking✦ Gamma Ray Bursts✦ SQL · Python · Tableau✦ BITS Pilani Hyderabad✦ IIT Bombay Thesis✦ Business Intelligence✦ Astronomy Club Lead✦ TEDx Speaker✦ Analytics✦ Product Thinking✦ Gamma Ray Bursts✦ SQL · Python · Tableau✦ BITS Pilani Hyderabad✦ IIT Bombay Thesis✦ Business Intelligence✦ Astronomy Club Lead✦ TEDx Speaker✦
Atreyi Dasgupta
MSc Physics
BITS Pilani · IIT Bombay
⭐ TEDx Speaker
About Me

A physicist who
became a builder

I graduated with an Integrated MSc in Physics from BITS Pilani Hyderabad and completed my Master's thesis at IIT Bombay, where I built a real-time Gamma Ray Burst detection pipeline for AstroSat satellite data.

Along the way, I realised the thing I love most is turning complex, noisy signal into clear insight. Whether that signal is coming from a neutron star or a business dashboard, the problem-solving muscle is the same. That's what brought me to data and product.

Toolkit

Skills & stack

🗄️
Data Engineering
SQL · CTEs · Window Functions
Subqueries · Joins · NULL Handling
SQLite · Data Cleaning
📊
Visualisation
Tableau Public · Dashboard Design
KPI Architecture · Treemaps
Matplotlib · Seaborn
🐍
Programming
Python · Pandas · NumPy
MCMC Modelling · Scipy
Statistical Analysis
🧠
Product Thinking
Frameworks · User Research
PRDs · Product Case Studies
Metrics & OKRs · Prioritisation
📈
Business Analytics
Unit Economics · LTV/CAC
Cohort Analysis · Funnel Metrics
Revenue Operations · Forecasting
🔭
Research Methods
Bayesian Inference · MCMC
Signal Processing · Hypothesis Testing
Algorithm Design · Scientific Writing
Selected Work

Built with code,
shipped with intent

01 — Featured

Real-Time GRB Detection Pipeline — AstroSat

Master's thesis at IIT Bombay. Designed a production-grade detection system for Gamma Ray Bursts using live satellite data from CZTI aboard AstroSat. Engineered four independent detection algorithms (N-Sigma, Top-N, CuSum, Sum-Threshold) — each addressing different statistical regimes. The system handles signal noise, false-positive suppression, and real-time throughput constraints, directly mirroring how AI systems balance precision vs. recall in high-stakes pipelines.

Python Signal Processing Algorithm Design AstroSat / CZTI Statistical Modelling
View Thesis Report ↗
🛰️
02

Netflix Content Strategy Dashboard

Full-stack analytics pipeline — raw CSV to production dashboard. Wrote complex SQL (CTEs, window functions, aggregations) against SQLite, processed in Python, and visualised in Tableau with KPI tiles, content growth timeline, country distribution, and a genre treemap. Built for business decision-making, not just aesthetics.

SQL Python · Pandas Tableau SQLite Business Intelligence
View Live Dashboard ↗
03 🤖 AI Agent

AI-Powered Job Hunt Automation Agent

Built an end-to-end agentic workflow using n8n that automates the job search pipeline — scraping listings, matching against a target role profile, drafting personalised outreach messages, and tracking application status. Demonstrates applied understanding of LLM orchestration, prompt chaining, and multi-step agentic logic in a real-world product context.

n8n Workflows LLM Orchestration Prompt Engineering AI Agents Automation
View Post ↗
04 📋 Product

Spotify Podcast Discovery — Product Case Study

End-to-end product case study on Spotify's podcast discovery problem. Defined user personas (Casual Commuter, Deep Diver, Lapsed Listener), mapped friction points, prioritised a feature roadmap, and designed two solutions: a Smart Discovery Feed and a Podcast Taste Graph. Includes success metrics, pitfall mitigation, and business impact framing.

Product Strategy PRD User Research Metrics & OKRs Roadmapping
View Case Study ↗
05 ⚛️ Quantum × ML

Galaxy Detection Using Quantum Machine Learning

Demonstrated galaxy classification at 94% accuracy using a hybrid quantum-classical model on NASA imagery. Divided images into 16×16 pixel patches, encoded via a Parameterized Quantum Circuit (PQC) chosen for high expressibility and entanglement. Trained with Cross-Entropy loss and L-BFGS optimisation, combining PyTorch and Qiskit Machine Learning.

Python Qiskit ML PyTorch Quantum Computing NASA Imagery
View on GitHub ↗
06

Neutron Star Afterglow Modelling

Krittika Astronomy Club project using MCMC with a broken power law model to fit neutron star afterglow light curves. Applied Bayesian parameter estimation to constrain physical decay properties post-merger. Implemented entirely from scratch in Python.

Python MCMC Bayesian Inference Astrophysics
View on GitHub ↗
Live Projects

Apps I built

📰
Live

TLDR

news, but make it readable.

AI-powered news app with 9 personality-driven channels — from "Trump Did What Now" to "Big Bang Bulletin". Click a channel, Groq summarises the latest headlines in that channel's Gen Z voice. Built with React, Vite, GNews API and Groq.

React Groq AI GNews API Vercel
Open App ↗
TLDR
🤡 Trump Did What Now
🚀 BuzzNews
🏏 Thala For A Reason
🎬 Bolly Gossip
🔭 Big Bang Bulletin
🎵
Live

ReadTheRoom

Wordle, but make it music.

Daily song-guessing game where you identify a track from increasingly obvious clues. Built in ~3.5 hours as a portfolio project. Uses React, Vite, Tailwind, Vercel serverless functions and Upstash Redis for daily state.

React Vite Tailwind Upstash Redis
Play Now ↗
🎵 ReadTheRoom
♪
♫
♩
♬
♭
Clue 2 of 5 unlocked
Experience

Where I've
done the work

2024 — 2025
Master's Thesis Researcher
Indian Institute of Technology Bombay
  • Designed and implemented a real-time Gamma Ray Burst detection pipeline for AstroSat satellite CZTI data
  • Built four detection algorithms from scratch: N-Sigma, Top-N, CuSum, Sum-Threshold
  • Translated scientific methodology into rigorous, reproducible code in Python
2023
Solar Flare Research
Remote Research Collaboration
  • Statistical classification of solar flare events using observational data
  • Applied supervised approaches to categorise flare intensity and duration profiles
  • Contributed to analysis pipeline and documentation
2023 — 2024
President, Ad Astra — Astronomy Club
BITS Pilani, Hyderabad Campus
  • Led a 50+ member club across technical workshops, observation events, and outreach initiatives
  • Organised a large-scale public event around the Chandrayaan-3 landing with campus-wide participation
  • Built the club's programming calendar and managed cross-team coordination throughout
2024
TEDx Speaker
TEDx BITS Pilani Hyderabad
  • Invited to speak at the campus TEDx event, presenting ideas at the intersection of success and hardwork, when life gets challenging.
Research Background

Where science
meets signal

💥
Gamma Ray Burst Detection
IIT Bombay · Master's Thesis · 2024-25
Real-time detection pipeline on AstroSat satellite data. Four algorithms designed to flag transient high-energy events in CZTI instrument data. Core challenge: maximising sensitivity while minimising false positive rate.
🌟
Solar Flare Classification
Remote Research · 2023
Statistical classification of solar flare events. Built models to categorise flares by intensity profile and temporal signature using observational light curve data.
⭐
Neutron Star Afterglow Modelling
Krittika Astronomy Club · IITB · 2023
MCMC fitting of a broken power law model to neutron star afterglow light curves. Bayesian parameter estimation for post-merger decay physics. Implemented from scratch in Python.
🎓
MSc Physics — Integrated
BITS Pilani, Hyderabad · 2021-2025
Four-year integrated programme covering quantum mechanics, statistical physics, electrodynamics, computational methods, and astrophysics. Strong grounding in Bayesian reasoning and analytical problem-solving.
A little more about me
"I spent years studying how the universe communicates through light curves and signals. Now I want to help businesses do the same with data."

Physics taught me to start with priors, update on evidence, and never mistake noise for signal. That's exactly how I approach product decisions, business problems, and analytical work. The methods scale.

Get In Touch
Let's build
something good
together.
✉️
Email atreyd03@gmail.com
→
💼
LinkedIn atreyi-dasgupta
→
📊
Tableau Public Netflix Dashboard
→
⚡
GitHub atreyidasgupta
→