const engineer = "deepak"; import { react, fastapi } from "skills"; ๐Ÿค– AI-powered apps
Available for hire ยท Graduating June 2026

Deepak Sai Full-Stack Engineer

Building AI-powered web applications that solve real problems. Final-year CSE student at VIT Chennai โ€” passionate about backend systems, RAG pipelines, and elegant UI.

3+
AI Projects
150+
LeetCode Solves
7.71
CGPA
3+
Certifications
Scroll to explore
01

About Me

I'm a final-year B.Tech CSE student at VIT Chennai who loves building things that work beautifully. My focus is on full-stack development and AI-powered tools โ€” from designing REST APIs to deploying RAG-based chatbots.


I work primarily with Python, React, and FastAPI, and I'm fascinated by how LangChain, vector databases, and LLMs can be combined into production-grade applications. Every project I build, I aim for clean architecture, real performance, and a UI that feels right.

๐Ÿง 
3+
AI Projects Built
โšก
150+
DSA Problems
๐ŸŽ“
7.71
CGPA at VIT
๐Ÿ…
3+
Certifications
02

Projects

Project 01
โ†—

QueryPDF Chatbot

A production-ready Retrieval-Augmented Generation chatbot that lets users chat naturally with multiple PDFs simultaneously โ€” powered by Google Gemini embeddings and Groq's blazing-fast LLM inference.

  • Multi-PDF ingestion with sub-second semantic search via FAISS vector store and Google Gemini embeddings, enabling citation-based natural language answers.
  • Full-stack architecture: Flask REST backend + React frontend with side-by-side PDF viewer and real-time chat interface.
  • Groq API integration for low-latency LLM responses across large document sets without degrading retrieval accuracy.
Project 02
โ†—

Budget Planner

A secure, full-featured financial management application with real-time analytics dashboards, helping users track and understand their spending across multiple categories.

  • JWT-based authentication with CRUD transaction management and a real-time spending analytics dashboard with monthly breakdown visualizations.
  • Interactive Recharts-powered charts surface spending insights with multi-category trend tracking at a glance.
  • Scalable FastAPI backend with MongoDB aggregation pipelines supporting pagination and category-level filtering.
Project 03
โ†—

Real-Time Waste Monitoring System

An end-to-end deep learning application that classifies waste into 9 categories in real time using a fine-tuned MobileNetV2 model, deployed on Streamlit Cloud with live camera capture.

  • Fine-tuned MobileNetV2 to classify waste (plastic, glass, metal, paper, etc.) with real-time confidence scoring deployed live on Streamlit Cloud.
  • Live camera detection + image upload with SQLite database for base64-encoded image persistence and long-term analytics.
  • Multi-page Streamlit UI with interactive charts, history filtering, and record management for real-time monitoring.
03

Tech Stack

๐Ÿ’ป
Languages
Python JavaScript C++ C Java
๐ŸŽจ
Frontend
React HTML5 CSS3 Tailwind CSS Recharts
โš™๏ธ
Backend
FastAPI Flask REST APIs JWT Auth
๐Ÿค–
AI / ML
LangChain TensorFlow FAISS Gemini API Groq API Pandas NumPy
๐Ÿ—„๏ธ
Databases
MongoDB SQL SQLite
๐Ÿ› ๏ธ
Tools & Cloud
Git & GitHub VS Code Postman Azure AI-900 Streamlit
04

Education

๐ŸŽ“
Vellore Institute of Technology, Chennai
B.Tech in Computer Science & Engineering
7.71 CGPA
Sept 2022 โ€“ June 2026
๐Ÿ“š
Sri Chaitanya Junior College
Intermediate (CBSE) ยท Visakhapatnam
77.4%
June 2020 โ€“ May 2022
๐Ÿซ
Dr. KKR Gowtham School
High School (CBSE) ยท Visakhapatnam
82.8%
June 2019 โ€“ March 2020
05

Certifications

Microsoft
Azure AI Fundamentals (AI-900)
Cisco Networking Academy
Introduction to Cyber Security
Coursera
SQL for Data Science
Let's Work Together

Got a project?
Let's build it.

Open to full-stack, backend, or AI engineering roles. Let's connect.