Software Developer · Backend Engineer · AI Explorer
I build practical software, backend systems, and AI-powered applications — turning ideas into useful products.
Learn → Build → Debug → Improve
I started with Python and expanded into backend development, APIs, databases, Java, and Spring Boot. I learn primarily by building projects, experimenting with technologies, debugging problems, and improving through iteration.
View Jay's Projects & ExperienceLEARN
Python, CS & Fundamentals
BUILD
RegiNova AI & Life OS
DEBUG
System Optimization & Fixes
IMPROVE
Java, Spring & Microservices
// Java & Spring Boot RAG Engine
@Service
public class VectorRAGService {
@Autowired
private FAISSVectorStore store;
public List<Document> query() {
return store.similaritySearch();
}
}
Python, FastAPI, REST APIs, PostgreSQL, SQL, Git & GitHub for robust production systems.
Java, Spring Boot, Node.js, TypeScript & Retrieval-Augmented Generation (RAG) architectures.
AI-Assisted Development, LLM Applications, Vector Search, Embedded Systems & Backend Architecture.
A platform making government documents easier to search, retrieve and understand through AI document intelligence, FAISS vector search, and LLaMA/Groq LLM RAG.
Interactive portfolio architecture featuring Next.js 14, Framer Motion scroll mechanics, WebGL canvas effects, and dark mode UI design system.
Industrial web application and microservices backend system engineered for real-time tracking, resource management, and service coordination.
Computer vision pipeline for automatic vehicle detection, license plate region extraction, preprocessing, and OCR text extraction.
Backend API system for storing digital content and releasing access at future timestamps using AWS S3 storage, Cron jobs, and JWT auth.
Machine learning system forecasting test-case failures from execution history data and engineered quality metrics.
Discover how Jay Tavanoji engineered government document intelligence using FAISS 384-dimensional vector indexing, FastAPI microservices, PostgreSQL, and LLaMA 3.3 RAG.
Jay Tavanoji's flagship AI platform for government document analysis, RAG pipeline automation, and sub-20ms FAISS vector searches.
High-throughput semantic search benchmarking FAISS vector indexing across 100,000+ legal & government documents with dense embeddings.
Real-time LLM token stream benchmark utilizing Groq LPU hardware acceleration for ultra-low latency contextual response synthesis.
Real-time vehicle license plate detection and optical character recognition pipeline processing HD camera streams with low overhead.
Event-driven cloud architecture utilizing AWS Lambda, DynamoDB, and S3 for scalable zero-idle cost backend automation.
