developer with strong fundamentals and having experience of building things that scales.

leetcode

github

Lakhman Parmar  • Peerlist

ArogyaSetu is an AI-driven healthcare assistant accessible through WhatsApp, designed to bridge India's healthcare accessibility gap — especially for rural and elderly populations — by offering real-time medical guidance in regional languages without needing an app. Key Features: -Voice support (English) -AI-based symptom analysis and contextual health recommendations -Zero-cost access via WhatsApp — no app install needed Tech Stack: - Backend: Python, FastAPI - AI/ML: Whisper ASR, Large Language Models (LLMs) - Memory: ChromaDB - Security: JWT Authentication - DevOps: Docker - Database: SQLite - Integration: WhatsApp Cloud API Deployment: - Docker containerization - CI/CD pipeline with GitHub Actions - Deployed on Azure Container Instances Key Learnings: - WhatsApp API integration for real-time interaction - Contextual AI for healthcare guidance - Docker containerization and CI/CD pipelines - Deployment using Azure Cloud My Role: - Integrated WhatsApp Cloud API for real-time communication - Developed core FastAPI backend to process AI interactions Impact: - Built to serve the 70% of India's population without consistent healthcare access - Offers intuitive health support through a familiar interface (WhatsApp) - Scalable, containerized architecture ready for deployment at scale

ArogyaSetu – AI-Powered Healthcare Companion

Full-stack MERN application for educators to create virtual classrooms with custom video processing pipeline. Features -   Upload videos → automatic conversion to multiple resolutions (ABS using FFMPEG/HLS) -   Virtual classroom management with shareable codes -   JWT authentication -   Job queue for video processing -   Take notes during lectures -   Comment,Like Lectures -   Adaptive bitrate streaming (.m3u8 segmentation) Tech Stack -   Frontend: React -   Backend: Node.js, Express -   Database: MongoDB -   DevOps: Docker, GitHub Actions, Azure Container Apps -   Video Processing: FFMPEG, HLS Deployment -   CI/CD pipeline with GitHub Actions -   Docker containerization -   Deployed on Azure Container Instances Key Learnings -   Adaptive bitrate streaming implementation -   Video processing pipelines -   Docker containerization -   CI/CD automation -   Cloud deployment strategies

Sarthi – Adaptive Video Streaming Classroom

Full-Stack Engineer | 2.5 months | Surat-based textile manufacturer Engineered a full-stack IoT monitoring platform for real-time loom machine tracking across multiple textile factories, processing 1M+ daily machine logs on a single server. Key Achievements: • Designed multi-tenant architecture with complete database isolation per factory, enabling independent schema migrations and automated provisioning • Built high-throughput data pipeline using TimescaleDB and MQTT, handling continuous telemetry from 500+ loom machines with offline queue support • Developed real-time shift analytics engine with PostgreSQL functions for day/night shift calculations spanning midnight boundaries • Implemented automated WhatsApp report delivery using BullMQ job queues with shift-completion detection • Created comprehensive dashboards with live machine status, efficiency heatmaps, and PDF/Excel export capabilities Tech Stack: Node.js, Express, PostgreSQL, TimescaleDB, React, TypeScript, Electron, MQTT, BullMQ, Redis Features: Real-time monitoring, automated shift reports, operational analytics, performance tracking, beam flow management, yarn inventory, admin panel with tenant management