I'm a final-year Electronics & Computer Engineering student at PICT, Pune (CGPA: 9.35/10.0), building scalable full-stack applications and AI-powered systems β from real-time trading platforms to deepfake detection pipelines.
- π¬ Currently working on a multimodal audio-visual deepfake detection project, targeting an IEEE journal publication.
- π» Building with the MERN stack, PyTorch, and LLMs (Google Gemini, LLaMA 3.3).
- π Competitive Programming: 2β CodeChef (Max Rating 1469) Β· Max Rating 1615 on LeetCode Β· 500+ DSA problems solved across LeetCode, CodeChef & GeeksforGeeks.
- π Hackathons: Finalist β SciTech Innovation Hackathon 2025 (Top 15/250+ teams) & Ignition HackVerse 2026 (Top 6/450+ teams).
- π Certified in Data Analytics Level 1 (MetricMinds Γ PICT).
- π Check out my portfolio website to see my work in one place!
π‘οΈ DeepShield β PyTorch FastAPI React.js LLaMA 3.3 MTCNN ViT
A deepfake detection system processing image, video, and audio inputs β combining MTCNN, Vision Transformers, and SigLIP for multimedia manipulation detection, with LLaMA 3.3 (via Groq API) generating explainable AI reports.
π TradeX β Node.js Express.js React.js MongoDB Socket.io JWT
A full-stack stock trading platform with live market tracking and portfolio management. Integrated the Yahoo Finance API over Socket.io WebSockets for real-time price streaming (no HTTP polling), with JWT auth, bcrypt hashing, and Joi-validated REST APIs.
π DataMart β Python Flask React.js PostgreSQL Pandas Scikit-learn Groq API
An AI-powered Business Intelligence platform with ETL pipelines, OLTP/OLAP processing, and a Snowflake-schema data warehouse. Interactive dashboards for customer segmentation, sales forecasting, and market-basket analysis, plus Groq-LLM-generated business insights.
π€ NoteBot β Python Streamlit LangChain Google Gemini NumPy
A RAG-powered study assistant for PDF-based Q&A, AI summaries, and flashcards β using vector embeddings, cosine similarity, and LangChain document chunking for accurate semantic retrieval.