Case Studies

Featured Projects

Scalable platforms built with AI, analytics, and intelligent automation.

Inboxonic – AI-Powered Email Intelligence & Automation System

Solo

Inboxonic is a full-stack AI-driven email automation system deeply integrated with Gmail. It uses a custom Gemini-powered NLP pipeline to summarize emails, detect intent, classify urgency, and generate intelligent reply drafts. The platform automates inbox workflows—including continuous email fetching, thread detection, classification, and event logging—and provides a real-time dashboard for analytics. Built with a distributed microservice architecture using Node.js, FastAPI, MongoDB, and secure OAuth2/JWT flows.

Tech Stack

Node.jsFastAPIGmail APIGemini AIMongoDB

Impact

  • Designed a distributed microservice architecture for high reliability
  • Implemented custom NLP pipeline for summarization, tone & intent detection
  • Built real-time dashboard powered by DailyMetrics aggregation engine
  • Developed secure OAuth2 integration with encrypted refresh tokens
  • Optimized MongoDB event logs for fast filtering & analytics

MedQuery Agent – AI Medical RAG Assistant

2 devs

MedQuery is a multi-role healthcare intelligence platform built for Doctors, Researchers, Patients, and Admins. It delivers fast, accurate, and medically-grounded answers using Retrieval-Augmented Generation, combined with a Doc Reading AI Agent capable of understanding PDFs, medical reports, and research papers. The platform supports patient report uploads, dataset analysis, research insights, and medical query history — powered by a FastAPI backend, FAISS vector search, and NeonDB metadata storage. Deployed on Google Cloud Run with complete Dockerization.

Tech Stack

React (TypeScript)FastAPIRAG PipelineFAISSOpenAI API

Impact

  • Built a fully automated medical document ingestion & embedding pipeline
  • Delivered accurate, source-cited medical insights using RAG
  • Optimized FAISS retrieval for high-volume medical data
  • Successfully deployed scalable backend on Google Cloud Run