Skip to content
View sm5190's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report sm5190

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
sm5190/README.md

👋 Hi there, I’m Shutonu Mitra (@sm5190)

I’m a backend-leaning Software Engineer with experience building scalable APIs, cloud-native services, and data-intensive systems, and integrating applied AI into production workflows.
My work focuses on reliability, performance, and clean system design at scale.

🔗 Connect


🧠 What I’m Currently Working On

  • 🧓 Fall Detection Platform
    Building a privacy-preserving, edge-to-cloud backend for indoor fall detection using mmWave radar, with real-time inference and alerting pipelines.

  • 🤖 Robotics & RL Systems (MuJoCo)
    Designing reinforcement learning agents with stability and retry metrics, focusing on evaluation, deployment readiness, and system robustness.

  • 🧩 Digital Twin Systems
    Developing backend services and data pipelines to simulate and monitor real-world systems using streaming data and cloud infrastructure.


🚀 Notable Projects

TypeScript | LangChain | Serverless | LLMs

  • Built an AI agent to assist with job applications and resume customization
  • Designed multi-agent pipelines and scalable backend services for feedback generation
  • Focused on reliability, prompt versioning, and production deployment

Python | FastAPI | HTTPX | Data Normalization | Docker

  • Built a resilient scraper to fetch real-time lines and player props from FanDuel, DraftKings, and PrizePicks
  • Designed a normalized odds schema to unify heterogeneous sportsbook APIs into a single event market selection model
  • Implemented rate-limiting, retries, freshness checks, and provider health tracking for production reliability

TypeScript | Gemini API

  • Built a system that converts research papers into executable algorithm implementations
  • Designed backend workflows for document parsing, reasoning, and code generation

React | Java | Tomcat | MySQL | AWS

  • Developed an end-to-end e-commerce system with REST APIs and SPA frontend
  • Optimized backend queries and connection pooling to reduce latency under load
  • Deployed on AWS with VPC isolation and secure public/private subnet architecture

Java | AWS Lambda | API Gateway | RDS Proxy | Terraform

  • Architected a serverless backend supporting burst traffic with low operational overhead
  • Implemented retries, DLQs, and automated IaC deployments for resilience

Android | IoT | ML | Firebase

  • Built an end-to-end system streaming encrypted biosensor data from mobile to cloud
  • Designed backend validation, storage, and analytics pipelines

PyTorch | TensorFlow | NLP

  • Designed a real-time phishing and malware URL detection backend
  • Achieved high accuracy using deep learning–based text representations

PySpark | AWS EMR | S3

  • Built a distributed MapReduce pipeline to analyze large-scale Twitter data
  • Processed and ranked trending hashtags using Spark on AWS

🧰 Tech Stack

Languages: Java, Python, TypeScript, JavaScript, SQL, Bash
Backend: REST APIs, Microservices, Spring Boot, Node.js, FastAPI
Cloud & Infra: AWS (EC2, Lambda, RDS, S3, EMR), GCP, Docker, Kubernetes, Terraform
Data & Messaging: MySQL, PostgreSQL, DynamoDB, Redis, Spark, Airflow, Kafka
AI & ML: LangChain, Hugging Face, PyTorch, scikit-learn
Frontend (when needed): React, Next.js, Tailwind CSS



I enjoy turning complex requirements into reliable systems — building software that scales, integrates intelligence thoughtfully, and holds up in production.

Pinned Loading

  1. Cyberbully-Detection Cyberbully-Detection Public

    Undergrad Thesis in the domain of NLP, Social Media and Machine Learning

    Jupyter Notebook 2

  2. GAS-Leakage-MPMC GAS-Leakage-MPMC Public

    An IoT based system capable of detecting gas leaks and sending alert message via SMS to the user's cellphone as well as triggering an alarm.

    C++ 1

  3. Predictis Predictis Public

    A medical decision making diagnosis system, which is an Android Application that predicts its user’s future possibility of having Cardio-vascular diseases using Machine Learning, Data Analytics and…

    Jupyter Notebook 2

  4. Malicious-URL-Detection Malicious-URL-Detection Public

    Detection of phishing, malware and defacement by character-based CNN

    Jupyter Notebook 1

  5. PaperTown-Book-store PaperTown-Book-store Public

    An E-Comic Bookstore in React and Rest API

    TypeScript

  6. Pyspark--hashtag-counter Pyspark--hashtag-counter Public

    Utilized Amazon EMR and Apache Spark for efficient data processing, the project features custom parallel processing code in Python to implement a map-reduce strategy.

    Python 1