Vansh Yadav

product manager

Indian Institute of Technology Roorkeefull_time, internship, freelance

Open to roles

Product Roadmap

Queue Cure '26 — Real-time Clinic Queue Management

Millions of clinics in India still use paper token slips and manual calling. Patients wait 2-3 hours with zero visibility into their position. Receptionists manage everything from memory — no dashboard, no estimated wait time, no system. Process Started by identifying the core problem: not long queues, but information asymmetry. Patients have no data; receptionists have all of it. Chose WebSockets over polling immediately — the live sync requirement needed push-based updates, not repeated fetching. Built frontend first with mock data, then replaced every state mutation with socket events. Deployed frontend on Vercel, backend on Render as a persistent Node.js server. What didn't work: tried deploying both on Vercel — WebSockets require a persistent server, which Vercel doesn't support. Switched backend to Render. Results Both screens update in under 100ms when "Call Next" is clicked — no refresh, no polling. Wait time calculates dynamically from real queue data. Concurrency lock prevents duplicate advances on rapid clicks. Patient reconnecting mid-session immediately sees correct state via REST hydration + on-connect socket emit. Reflection Add Redis for state persistence so a server restart doesn't reset the queue. Add PIN-based auth on the receptionist screen. Replace manual token entry with SMS delivery so patients don't need their paper slip to track their wait.

4 media files · queue-2026-haxa.vercel.appView

IPL Crunch '26 — What Actually Wins IPL Matches: 19 Seasons, 1,218 Matches, Data-Backed

Most IPL analysis online only talks about basic things like win/loss records or top scorers. I wanted to go deeper and find patterns that fans usually don’t notice. Using ball-by-ball data from 19 IPL seasons covering 1,218 matches and nearly 2.9 lakh deliveries, I explored questions like: How much do powerplay wickets actually affect winning? Is chasing really an advantage? Which bowlers perform best in death overs? At what point does momentum usually shift in a match? Process I started by cleaning and organizing the IPL dataset using Python and pandas. After understanding the structure of the data, I focused on four main areas: powerplay wickets, chasing trends, death-over bowling, and momentum shifts during matches. I used matplotlib and seaborn to create charts that made the insights easier to understand. I also wrote a separate verification script to cross-check all calculations. One challenge I faced was while filtering death-over bowlers. My first criteria were too strict, which removed many important players from the analysis. After adjusting the conditions, I got a much more balanced and useful dataset. Results The analysis revealed several useful insights. Teams losing only 0–1 wicket in the powerplay had a much higher win rate compared to teams losing 2 or more wickets. Chasing teams also won slightly more matches overall. Death overs turned out to be the highest-scoring phase of the game, while bowlers like DJ Bravo, Bhuvneshwar Kumar, Lasith Malinga, and Jasprit Bumrah stood out as strong death-over performers. I also noticed that average IPL scores have steadily increased over the years, showing how aggressive batting has evolved in modern T20 cricket. Reflection If I worked on this project again, I would: Add player-level pressure analysis for clutch situations. Compare team-specific strategies while chasing or defending totals. Use interactive tools like Plotly for better data exploration. Include stadium and pitch impact on scoring patterns. Track player performance trends across multiple seasons.

11 media filesView

This is Vansh’s work on Wooble.

Build a profile that shows what you can do — and share it anywhere.

Build yours