I’m a CS undergrad at MNIT Jaipur (2022–2026) and I build production-style web apps (MERN) and practical ML systems that run beyond notebooks.

• I build role-based, JWT-secured web platforms with React + Node.js + Express + MongoDB, including validation, rate-limiting and clean API boundaries (see: Laundry Management System).
• I’ve implemented full-stack dashboards with charts and transaction pipelines (see: Personal Budget Tracker with Recharts + MERN backend).
• I work on computer vision / segmentation systems designed to run on CPU, including TorchScript export and dynamic INT8 quantization for inference (see: land_seg).
• I’ve shipped an ML project that connects Apriori association rules with prediction models and a Streamlit UI (see: Smart_Retail_System).
• I’m open to SDE / Backend (Node.js) and ML / Computer Vision roles — anywhere / remote.

Only includes technologies that appear in your repositories (READMEs + language breakdowns).

• What it does: Booking workflow + status tracking, dashboards, admin analytics, pricing, and secure access separation.
• Tech: React 18, Vite, Tailwind CSS, Node.js, Express, MongoDB (Mongoose), JWT, bcrypt, express-validator, rate-limiting
• Repo: https://github.com/Bhavdeepmnit/Laundry_Management_System

• What it does: Tracks income/expenses with charts, authentication, transaction history, and export to Excel.
• Tech: React, Tailwind CSS, Recharts, Node.js, Express, MongoDB, JWT, bcrypt
• Repo: https://github.com/Bhavdeepmnit/Personal_Budget_Tracker

• What it does: Semantic segmentation pipeline designed for CPU-only training/inference, with TorchScript export and dynamic INT8 quantization options.
• Tech: Python, PyTorch, U-Net (MobileNetV3 encoder option), ONNX (optional), export tooling
• Repo: https://github.com/Bhavdeepmnit/land_seg

• Hardening MERN backends: JWT auth flows, input validation, role-based access, and rate-limited auth routes (patterns used in Laundry_Management_System).
• Deployment-ready ML pipelines: reproducible training + evaluation + export (TorchScript) / quantization (INT8) for CPU inference (from land_seg).
• TypeScript-first apps: using TS-heavy repos (e.g., CaseCobra, EduAssistant) to move more frontend code to typed, scalable patterns.

_{Built from real projects in this account: MERN platforms, ML pipelines, and TypeScript apps — always aiming for clean architecture and deployable outcomes.}