sameer Kumar mishra

Data scientist

fresherGandhi Institute of Engineering and Technology University, Odisha, Gunupur (giet University)full_time, internship

Open to roles

PythonSQLMachine Learningpandas

Ambula — Full Stack Healthcare Booking Platform

Millions of patients in India still find doctors through phone calls and paper registers. Clinics manage appointments on WhatsApp. Doctors have no digital view of their day. Patients carry physical files between consultations. I set out to build a platform that fixes this for both sides — patients can search doctors, view real slots, and confirm bookings in under 2 minutes on mobile. Doctors get a dashboard showing today's schedule, patient health history before the consultation begins, and tools to add diagnosis, prescription, and notes after each visit. The hardest constraint: if tw Process Started by mapping the two user journeys separately — patient flow and doctor flow — before writing a single line of code. Identified the hardest technical problem first: concurrent booking. Solved it at the database level using MongoDB's atomic findOneAndUpdate with a conditional filter, so the slot claim and status flip happen in one operation with no race window. Built backend first — Express REST API with five route modules, JWT auth for doctors and patients separately, and Mongoose models with compound indexes for fast slot queries. Seeded 8 doctors with 7 days of 30-minute slots each. Frontend built in React with React Router, a global CSS design system, and zero component libraries — everything custom. Mobile-first layout tested at 375px. Added analytics charts using pur Results Delivered a fully working platform in 14 days with both patient and doctor sides complete. Patients can search doctors by specialization and city, view 7 days of available slots, and book in under 2 minutes on mobile. Doctors get a dashboard with today's schedule, patient health summary before each consultation, and tools to add diagnosis and prescription after each visit. Double-booking prevention works correctly under concurrent requests. Email confirmations send automatically on booking. Prescription PDF downloads with doctor signature. Analytics dashboard shows revenue, patient co Reflection I would add Razorpay payment integration before the booking confirmation so consultation fees are collected online — this removes the cash dependency at the clinic and gives the platform a revenue model. I would replace polling-based slot availability with WebSockets so slots grey out in real time across all open browsers the moment someone books them — currently a patient sees a slot as available until they try to book it. I would also add a proper test suite using Jest and Supertest covering the atomic booking route with simulated concurrent requests, so the double-booking guarantee

8 media files · ambula-cyan.vercel.appView

IPL CRUNCH '26 — 5-Season Ball-by-Ball Analysis

Everyone has IPL opinions — "toss is everything", "death overs win matches", "Kohli always delivers Process Started with raw ball-by-ball CSV data where every row was one delivery bowled. Cleaned and validat Results Three questions answered with data: 1. Toss — Toss winners win only 46.7% of matches. Toss losers a Reflection I'd add economy rate and strike rate to the bowler and batter rankings rather than raw wickets and

8 media files · sameermishra811466.github.ioView

This is sameer’s work on Wooble.

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

Build yours