Engineering Experience

I’m Rupankar Dutta, a B.Tech student and builder with hands-on experience across embedded systems, Edge AI, backend development, and full-stack web applications. My work is centered on practical systems that combine hardware, software, and real-time decision-making.

Focus Areas

Embedded Systems, IoT, and Edge AI

  • Built decentralized real-time systems on ESP32 for local communication, sensing, and automation
  • Developed firmware for sensor-integrated devices using MPU-6050, DHT11, MQ-135, PIR, and sound sensors
  • Worked with ESP32-WROOM-32, ESP32-CAM, and ESP-NOW for low-latency wireless communication
  • Trained and deployed on-device ML models using TinyML, LSTM autoencoders, int8 quantization, TFLite Micro, Edge Impulse, and MobileNetV2

Backend, Full-Stack, and AI Application Development

  • Built application backends using FastAPI, Flask, Node.js, and Express
  • Engineered microservices infrastructure including API gateway patterns with JWT authentication and Redis rate limiting
  • Developed interactive web experiences using React, Socket.IO, MySQL, and PostgreSQL
  • Integrated AI services (OpenAI API, Google Vision) into full-stack products with personalization and chat memory
  • Containerized applications with Docker and set up CI workflows with GitHub Actions

Programming and Tooling

  • Languages: Python, C, C++, JavaScript, SQL
  • Tools: Arduino IDE, PlatformIO, EasyEDA, Git, Docker, VS Code
  • Libraries and utilities: BeautifulSoup, PyPDF2

Project Experience

Edge AI Anomaly Detection System (GitHub)

  • Built a real-time vibration anomaly detection system running entirely on an ESP32 with an MPU-6050 sensor — no cloud required
  • Trained an LSTM autoencoder in TensorFlow on normal-only data and quantized it to int8 for TFLite Micro on-device inference
  • Added a fault-type classifier that identifies the kind of anomaly (drop, shake, imbalance) alongside severity
  • Developed a Flask + JavaScript dashboard with live reconstruction-error charts, adaptive EWMA thresholding, and CSV-exportable anomaly event logs

CookMind AI – AI-Powered Cooking Assistant (GitHub)

  • Developed a full-stack AI application using React, Vite, FastAPI, PostgreSQL, and the OpenAI API
  • Built recipe recommendations from ingredient lists, conversational cooking assistance with chat memory, and image-based ingredient recognition via Google Vision
  • Implemented JWT signup/login, pantry persistence, cuisine and dietary preferences, and favorite recipe saving
  • Containerized the stack with Docker (dev and production compose files with an Nginx reverse proxy) and added a CI workflow for backend tests and frontend builds

TheRUPOgate – Microservices API Gateway (GitHub)

  • Engineered a production-style Node.js API gateway serving as a single entry point in front of microservices
  • Implemented JWT authentication with role-based header injection for downstream services
  • Built Redis sliding-window rate limiting that fails open if Redis is unavailable
  • Added structured JSON logging and a live monitoring dashboard streaming request metrics over Server-Sent Events

Intelligent Decentralized Home Automation System

  • Developed a decentralized smart home system using ESP32 nodes communicating through ESP-NOW
  • Implemented local sensing, threshold-based alerts, timing control, and offline automation logic
  • Integrated DHT11, MQ-135, PIR, and sound sensors into the firmware pipeline
  • Contributed to a MobileNetV2-based facial recognition workflow for secure access control on ESP32-CAM

Real-Time Multiplayer Quiz Platform

  • Built a multiplayer quiz application using React, Node.js, and Socket.IO
  • Implemented private rooms, timed rounds, live score tracking, leaderboard generation, and synchronized gameplay state

Natural Language to SQL Query Generator

  • Built a Flask and MySQL based application that converts natural language questions into SQL queries
  • Developed the backend query flow, execution pipeline, and result presentation interface

Dynamic Programming Text Auto-Correction

  • Developed a C-based text correction system using edit-distance dynamic programming
  • Focused on optimized string matching, fast suggestion generation, and correction accuracy

Automated Plant Watering System

  • Built an ESP32-based irrigation system that automates watering decisions from environmental input
  • Added real-time monitoring and control behavior to improve reliability and efficiency

WiFi Network Scanner

  • Developed an ESP32 utility to detect nearby WiFi networks and display SSID and RSSI values
  • Used the system for quick signal analysis and basic wireless environment diagnostics

For more details about my technical background and projects, please refer to my resume.