M.Tech Software Engineering @ VJTI Mumbai · Building at the intersection of full-stack web and applied AI.
- GenAI applications — RAG pipelines, LLM-grounded QA systems, embedding visualizations
- Deep learning systems — dual-task neural networks, XAI with LIME & Grad-CAM, NLP with TinyBERT
- Full-stack web apps — MERN stack, Node.js/Express, MongoDB, REST APIs
| Project | What it does | Stack |
|---|---|---|
| Swiggy RAG AI | Ask questions directly from Swiggy's annual report — RAG pipeline with semantic search, citation highlighting, conversation memory | Python · LangChain · FAISS · Flan-T5 · Streamlit |
| BOTDETECT | Twitter bot detection — 93.25% accuracy across 279k tweets using TinyBERT + DNN with AdaBelief optimizer and LIME explainability | PyTorch · TinyBERT · Optuna · LIME |
| MTech Leaf Disease | Dual-task EfficientNet: simultaneous disease classification + severity regression (0–100%) with Grad-CAM explanations | PyTorch · EfficientNet-B0 · OpenCV · Grad-CAM |
| Veranda | Airbnb-style village stay listing platform with full CRUD, RESTful routing, and server-side validation | Node.js · Express · MongoDB · EJS |
Python PyTorch TensorFlow LangChain RAG FAISS HuggingFace
React.js Node.js Express.js MongoDB MySQL REST APIs
Explainable AI NLP Computer Vision Optuna Streamlit
- M.Tech research: plant disease detection with interpretable deep learning
- Open to: AI/ML Engineer · Full-Stack Developer · GenAI Application roles
📍 Mumbai · LinkedIn · mishra.khushboo27093@gmail.com