👋 Hi, I'm Somasree Majumder
Full Stack Machine Learning Engineer | AI/ML Specialist | Open Source Contributor
Building intelligent systems that solve real-world problems through cutting-edge AI/ML technologies
🚀 About Me
I'm a Full Stack Machine Learning Engineer at DeepEdge, specializing in computer vision, deep learning, and scalable AI solutions. With a strong foundation in both research and production-grade ML systems, I bring ideas from paper to deployment.
- 🎯 Currently working on anti-UAV quadcopter tracking systems with advanced object detection and tracking
- 🔬 Published researcher with papers in Homomorphic Encryption and ML in Medicine
- 🏆 Top 10 finalist at Challang Hackathon (100 elite teams worldwide)
- 🌟 LiFT Scholar 2022 (Top 500 worldwide)
- 💡 Active contributor to PyTorch Lightning, Hugging Face, KerasCV, and KerasNLP
- 📊 Competitive programmer: Ranked 18/2500 in HackerEarth ML Challenge
- 🎓 B.Tech in Electronics and Communication Engineering with 8.94 CGPA
💼 Professional Experience
🔹 Full Stack Machine Learning Engineer | DeepEdge
March 2025 – Present | Hyderabad, India
Building defense-grade computer vision systems:
- 🎯 Developed end-to-end anti-UAV quadcopter tracking system for runaway detection under challenging infrared conditions
- 🚀 Achieved 50%+ performance improvement over YOLOv8s through motion queue masks optimization
- 🔬 Implemented MOG2 Background Subtraction + SVM Classification achieving 30% performance gains
- 📈 Integrated advanced tracking algorithms: Lucas-Kanade, TAPNet (GoogleDeepmind), CoTracker (Facebook)
- 🎛️ Built sophisticated tracking models with YOLOv8, Kalman filters (linear & non-linear), and infrared thermal optimization
- 👥 Successfully managed 2-3 simultaneous client projects while maintaining high service quality
🔹 Machine Learning Engineer and Developer | Deloitte
July 2023 – March 2025 | Hyderabad, India
Delivered enterprise-grade AI solutions:
- 🤖 Architected multilingual RAG pipeline with FAISS vector database, FastAPI endpoints, and Docker containerization
- 🎬 Engineered Django-based movie recommendation system using content-based filtering, FunkSVD, and popularity algorithms
- 📊 Developed predictive model for customer purchase likelihood achieving 90% accuracy through advanced feature engineering
🔹 Machine Learning Developer | Unify.ai (Freelance)
Feb 2023 – May 2023 | Remote
Contributed to cross-framework ML infrastructure:
- 🔧 Implemented robust functions across TensorFlow, JAX, NumPy, PyTorch, PaddlePaddle increasing code coverage by 10%
- ✅ Led comprehensive unit testing with Hypothesis framework achieving 80% test coverage
- 🛠️ Developed Ivy OpConverters for TensorRT backend and mentored contributors through complex codebases
🛠️ Technical Arsenal
Languages
AI/ML Frameworks
Computer Vision & Tracking
- YOLOv8, Kalman Filters (Linear & Non-linear)
- Lucas-Kanade Tracking, TAPNet, CoTracker
- MOG2 Background Subtraction, Infrared Thermal Optimization
- Object Detection, Point Tracking, Motion Analysis
GenAI & LLMs
- RAG Pipelines, Vector Databases (FAISS, ChromaDB, Pinecone)
- DeepSeek-R1, LLama Models, Groq, Ollama
Backend & APIs
Databases
DevOps & Cloud
🎯 Featured Projects
🤖 AI Agent Flight Scanner
Tech Stack: Groq | Ollama | DeepSeek-R1 | LangChain | SQLite | ChromaDB | FastAPI
Built intelligent flight data analysis system using advanced RAG architecture with DeepSeek-R1-Distill-LLama-70B and LLama 3.2:3B models for comprehensive flight analytics.
🛍️ Amazon Product Classification
Tech Stack: PyTorch | Deep Learning | ALBERT Transformer
Engineered sophisticated feature engineering and preprocessing pipeline improving model accuracy by 10% using synthetic PyTorch datasets and ALBERT transformer for advanced multiclass classification.
🍕 Django Food Delivery Platform
Tech Stack: Django REST Framework | SQLite | JWT Authentication
Architected comprehensive full-stack food delivery platform featuring:
- Vendor management system & proximity-based search
- Payment integration & dynamic menu management
- Automated tax calculations & real-time order tracking
- Push notification system
🧠 ChildMind Anxiety Prediction
Tech Stack: TensorFlow | Gated RNNs | Deep Learning
Developed innovative predictive healthcare model using Gated RNNs to analyze children's physical activity patterns for early detection of problematic internet use, contributing to child mental health research.
🔍 Custom Vector Database
Tech Stack: Python | HNSW Graphs
Built high-performance lightweight vector database using Hierarchical Navigable Small World (HNSW) graphs for efficient similarity search and retrieval operations.
🏆 Achievements & Recognition
🎖️ Competitive Success
- 🥇 Top 10 Finalist - Challang Hackathon (100 elite teams worldwide)
- 🥈 Rank 18/2500 - HackerEarth ML Challenge
- 🥉 Rank 12/1000+ - Cipla Competition
- 🌍 1.6k Globally - Google Code Jam for Women
📚 Research Publications
🌟 Elite Programs
- 🎓 LiFT Scholar 2022 - Top 500 worldwide
- 👥 GHC OSD'23 Mentor - OpenSSF project mentor
- 🌐 Active Open Source Contributor - PyTorch Lightning, Hugging Face Transformers, KerasCV, KerasNLP
📜 Certifications
📊 GitHub Statistics
📈 Contribution Graph
💡 Problem Solving Profile
Expertise in: Data Structures, Algorithms, Competitive Programming, Problem Solving
📫 Let's Connect!
⚡ "The only way to do great work is to love what you do." - Steve Jobs
💻 Open to collaboration on ML/AI projects | 🌟 Always learning and growing