I'm a data analyst who doesn't stop at the insight. When I find a broken system — in a warehouse, a supply chain, a fridge — I build something to fix it. Sometimes that's a Python script. Sometimes it's a dashboard. Sometimes it's a physical device with a circuit board and an OLED screen.
My day job is data: client reporting, consumer intelligence, operations analytics.
Outside of that, I build products.
AI is my build partner. It handles the code I don't know. The ideas, the product logic, and the "does this actually solve anything?" question are mine.
Three real experiment case studies — each one caught something a first-glance answer would have missed: a hidden data confound, the wrong success metric, and a clean result double-checked for hidden bias.
Built to practice one thing: not trusting a result just because the p-value looks good.
Python · Statistics · Hypothesis Testing · Pandas
→ A-B-Testing-Case-Studies
A hardware fridge tracker — ESP32 + OLED screen, works offline, mounts on your fridge with magnets.
Built because I kept wasting food and money on things I forgot I had.
C++ · WiFi hotspot · Web interface · Persistent storage
→ freezer-robot
A PWA to track the 75 Hard challenge — daily tasks, streak badges, grace system, weekly reflections.
Built for myself. Single HTML file. No installs, no accounts.
HTML · CSS · JS · PWA
→ 75_HardDay
Automated risk monitor — live APIs, risk scoring, Google Sheets, Gmail alerts via GitHub Actions.
Built because supply chain failures are silent until they're expensive.
Python · Automation · Google Sheets API · GitHub Actions
→ Supply-Chain-Early-Warning
Analysed 43,739 real Amazon deliveries to surface hidden ghost delivery patterns.
Python · SQL · Looker Studio · Jupyter
→ amazon-ghost-delivery-analysis
Automated facilities task management — auto-escalation, live forms, email alerts, Looker Studio dashboard.
Python · Google Sheets API · HTML · Data Visualisation
→ SmartOps-Task-Platform
- I build for myself first, then see if others have the same problem
- I'd rather ship something small and real than design something big and hypothetical
- Logic first. Then build. Then make it look good.
- I use AI as a tool — the ideas, the problem framing, and the judgment are mine
- 🤖 Freezer Robot v2 — battery, wake button, better case
Data Analyst by day — Python, SQL, Power BI, turning messy data into decisions.
Product builder by instinct — if something frustrates me enough, I'll build a fix for it.
📍 India · 💼 LinkedIn · 📸 @just_nav.neet
Everything here is built by me — with AI as the engine, not the driver.