職人
職人/shokunin/
Noun

The Artisan; dedicating oneself to the mastery of their craft.

Jayesh Rajbhar

Jayesh Rajbhar

Full Stack Developer | AI Engineer | Builder

About

I am a Full Stack Developer and AI Engineer driven by a passion for building scalable, high-performance applications. From fine-tuning BERT models for legal document analysis to designing microservices architectures with Spring Boot, I enjoy solving complex engineering problems end to end. I specialize in React and Next.js on the frontend, Node.js and FastAPI on the backend, Java microservices with Spring Boot, and deep learning systems with PyTorch. My goal is simple: to build software that is robust, intelligent, and delightful to use.

Certifications

Course
  • Gained hands-on experience with AWS services including S3, EC2, and Lambda for scalable web applications.
  • Designed and implemented cloud-based solutions reducing infrastructure costs by 25%.
  • Completed certification requirements demonstrating proficiency in cloud architecture principles.
  • Developed secure file storage solutions using AWS S3 with signed URL generation.

Projects

AI / ML2026

Legal Contract Analyzer

AI-powered legal risk analysis platform featuring 4 fine-tuned BERT models for clause classification, risk scoring, unfair clause detection, and NER. It processes contracts end to end in under 60 seconds, uses a multi-task Legal-BERT architecture trained on LEDGAR across 100+ clause types, and integrates a switchable LLM layer with Groq or Ollama plus RAG-grounded clause explanations.

  • 4 fine-tuned BERT models processing contracts in under 60 seconds
  • Multi-task Legal-BERT trained on LEDGAR across 100+ clause types
Next.jsTypeScriptFastAPIPythonPyTorchBERTHuggingFace TransformersGroqOllamaRAGPyMuPDFGitHub ActionsVercel
Full Stack2025

CodeHealth AI

Production-grade AI observability platform analyzing 10+ repository health and code quality metrics. Built with a Next.js and React TypeScript frontend for analytics dashboards and multi-view repository insights, with Redis caching, BullMQ workers, Socket.IO streaming, SSR, edge caching, and optimized data fetching.

  • Tracks 10+ repository health and code quality metrics
  • Improved dashboard load time by about 40% with SSR and edge caching
ReactNext.js 15TypeScriptNode.jsD3.jsRechartsPostgreSQLRedisBullMQSocket.IOGitHub ActionsVercelAWS
Full Stack2024

Chattly

Real-time chat application featuring WebSocket-based messaging, JWT authentication, and presence indicators. Built with a React frontend and a Node.js plus Socket.IO backend, persisting messages in MongoDB.

  • Real-time WebSocket messaging with presence indicators
  • JWT-based authentication
ReactNode.jsSocket.IOMongoDBJWT
Java / Microservices2026

MicroServices-HMS

Microservices-based Hospital Management System using Spring Boot, Spring Cloud Eureka, JWT authentication, API Gateway, and a Next.js TypeScript frontend. It splits auth, patient, doctor, staff, discovery, and gateway routing into independent services, improving modularity and service separation over the earlier monolith.

  • 6 independent microservices for auth, patient, doctor, staff, discovery, and gateway
  • Spring Cloud Eureka service discovery plus API Gateway routing
JavaSpring BootSpring Cloud EurekaSpring SecurityAPI GatewayJWTMySQLNext.jsTypeScriptTailwind CSS

GitHub Activity

Skills

Frontend

React.js
Next.js
TypeScript
JavaScript
HTML5
CSS3
Tailwind CSS
Zustand
D3.js
Recharts

Backend

Node.js
Express.js
FastAPI
Python
Java
Spring Boot
Spring Cloud
Go
REST APIs

AI / ML

PyTorch
HuggingFace Transformers
LLM APIs
RAG
NLP
scikit-learn
OpenAI Whisper

Databases & Caching

PostgreSQL
MongoDB
MySQL
Supabase
Redis

Real-Time & Async

Socket.IO
BullMQ
WebSockets
Background Workers

DevOps & Cloud

GitHub Actions
Vercel
Heroku
AWS S3

Let's work together

Have a project in mind? Let's create something amazing.

Email Me