Open to full-time software roles

Hey There! I am Vaibhav Sengar.

Software Developer focused on full-stack development, Generative AI, and scalable AI-powered solutions

Nagpur, India GitHub LinkedIn

About

A Bit About Me

I am a Computer Science graduate with a strong full-stack foundation and a growing specialization in Generative AI and deep learning systems.

I work best at the intersection of software engineering and AI: designing clean system architecture, shaping model pipelines, and shipping complete applications that turn complex ideas into usable products. My recent work spans medical image analysis, Retrieval-Augmented Generation, secure role-based dashboards, and API-first backend systems.

Full-stack systems

React, Node.js, FastAPI, MySQL

AI specialization

RAG agents, ViT models, Gemini workflows

Product thinking

Secure RBAC, clean APIs, practical UX

Skills

Tools I Work With

Languages

PythonJavaScriptTypeScriptJava

Backend

Node.jsExpress.jsFastAPIFlaskRESTful APIs

Frontend

React.jsNext.jsReduxHTML5CSS3Tailwind CSSStreamlit

AI / Generative AI

RAGLLMsLangChainHugging FacePrompt EngineeringSemantic Search

Databases

MySQLPostgreSQLMongoDBVector Databases

DevOps & Tools

DockerGitHubGitPostman

Projects

Things I've Built

Each project is framed around practical outcomes: accuracy, retrieval quality, secure access, and maintainable architecture.

Deep Learning

DFU Detection System using Vision Transformer

A multimodal Diabetic Foot Ulcer detection system for wound image analysis, built to support early-stage clinical screening and hospital discovery.

92% accuracy
91% precision
93% sensitivity
Project Description
  • Compared ViT, CNN, and YOLO pipelines across imbalanced medical image data, with ViT producing the strongest sensitivity and precision.
  • Built a sub-second inference endpoint and a React interface for predictions with location-based hospital suggestions.
PyTorchVision TransformerFastAPIReactOpenStreetMapLeaflet

Generative AI

AI-Powered Multilingual Customer Support RAG Agent

An end-to-end RAG assistant that answers enterprise document questions with citation-backed, context-aware responses.

Citation-backed answers
Semantic search
Dockerized app
Project Description
  • Implemented document ingestion with recursive chunking, embeddings, vector retrieval, prompt templates, and conversation memory.
  • Designed a modular backend and Streamlit interface for PDF upload, conversational querying, and scalable query handling.
PythonLangChainGemini APIFAISSFastAPIStreamlitDocker

Full Stack

Internship Management System

A role-aware management platform for interns and mentors, created during internship work at Info Origin Technologies.

10+ REST APIs
RBAC security
MVC architecture
Project Description
  • Created dedicated Intern and Mentor login flows with access control to protect sensitive records and dashboard actions.
  • Tested API flows in Postman and structured the application for maintainability, reliability, and smoother data movement.
ReactNode.jsExpress.jsMySQLPostman

Experience

Where I've Worked

Former Intern

Info Origin Technologies Pvt Ltd

Built a full-stack internship management system with React, Node.js, Express.js, and MySQL. Delivered secure role-based access, scalable REST APIs, MVC structure, and reliable backend testing workflows.

Stack Used

React, Node.js, Express.js, MySQL

Security

Secure, role-based access control

Delivered

10+ REST APIs, MVC architecture

Education

What I Studied

Education

B.Tech, Computer Science and Engineering

G.H. Raisoni College of Engineering
2021 - 2025

Degree

B.Tech, Computer Science and Engineering

Duration

2021 - 2025

Institution

G.H. Raisoni College of Engineering

Let's Connect

I am open to full-time opportunities where I can build reliable software, learn fast, and contribute to teams solving meaningful technical problems.