I am a Computer Science Engineering Student (Expected Graduation: 2027) at Rajeev Gandhi Memorial College of Engineering and Technology with a CGPA of 8+. I specialize in building Intelligent Applications, combining AI/ML architectures (Computer Vision, NLP) with robust Backend Development using modern paradigms.
- 🚀 Current Focus: Quantitative Computer Vision modeling, REST API engineering, and lightweight Edge AI deployment.
- 💡 Strategy: Fast, end-to-end prototyping of production workflows from dataset tuning to web serving.
- ♟️ Strategics: District-level Chess Player, bringing structured problem-solving and tactical thinking to system architectures.
- Backend / DB / Deploy: Flask, Supabase, Firebase, SQL, Go, Vercel.
- Machine Learning: YOLOv8, Scikit-learn, XGBoost, LightGBM, Transfer Learning, PaddleOCR, Ollama.
| Domain | Proficiency | Details |
|---|---|---|
| Computer Vision | Advanced | YOLOv8 depth estimation, QR code vision, local frame processing, and OCR |
| Tabular ML | Intermediate | Ontological classification systems, XGBoost, and SHAP explainability |
| Generative AI | Practical | Oracle Certified GenAI Professional, chat orchestrations, and LLM completions |
Built with: Python | Firebase | React | QR vision
- Designed QR-based logistics tracking dashboard that reduces manual verification operations.
- Engineered role-based access control (RBAC) and automated inventory threshold alert triggers.
- Ranked in the Top 10 Finalists in a national hackathon among 40+ colleges.
Built with: Python | YOLOv8 | PaddleOCR | Ollama | CV
- Created a fully offline spatial helper detecting real-time objects and obstacle distances.
- Integrated PaddleOCR text reading with text-to-speech feedback for user navigation.
Built with: LightGBM | XGBoost | Python | Streamlit
- Formulated ontology-driven ML pipeline processing 100k+ clinical health records.
- Achieved ~90-97% classification accuracy backed by SHAP explainability matrices.
Built with: Python | Flask | Transfer Learning
- Engineered blood cell classification models utilizing transfer learning strategies.
- Served real-time prediction overlays through a clean Flask API web dashboard.
2025
- Built responsive UI components and web layouts using React and HTML/CSS.
- Integrated dynamic API endpoints to display real-time metrics feeds.
2025
- Developed, trained, and served CNN-based HematoVision models on Flask backends.
- Optimized and cataloged training datasets, managing validation check tests.
2025
- Designed functional weather dashboards and interactive data showcases using Flask APIs.
- Top 10 Finalist in National Hackathon among 40+ competing engineering colleges.
- Qualified for TCS CodeVita Round 2 (competitive programming).
- District-level Chess Player (Strategic pattern solving).
Learning: "Advanced Neural Net Quantization Techniques (ONNX / TensorRT)"
Building: "Decentralized spatial audio navigation systems"
Exploring: "Serverless FastAPI microservices on edge functions"
OpenTo: "Machine Learning, Full-Stack engineering, and Open Source contributions""Translating complex algorithmic paradigms into functional spatial computing solutions."