Shivam Prajapati

02 Technical Skills

What I build with

A production-tested stack spanning backend systems, AI/LLM integration, cloud infrastructure, and full-stack web development.

Languages

Primary languages across backend services, scripting, and frontend development.

Java

Python

TypeScript

JavaScript

Backend & APIs

Production-grade frameworks for REST, GraphQL, event-driven, and microservice architectures.

Spring Boot

Spring Security

FastAPI

Node.js

GraphQL REST APIs OAuth 2.0 Webhooks Microservices

Frontend

Modern component-based UI development with server-side rendering and responsive design.

React.js

Next.js

HTML5

CSS3

Tailwind

AI & LLMs

End-to-end AI application development — from embedding pipelines to agentic workflows and production LLM integration.

Claude API LLM Integration RAG Pipelines LangChain ChromaDB Vector Search Prompt Engineering SSE Streaming QLoRA/LoRA Fine-Tuning HuggingFace Multi-Agent Systems

Cloud & DevOps

Cloud-native deployment, containerization, and CI/CD pipeline automation at production scale.

AWS

Docker

Kubernetes

Terraform

Jenkins

Git CI/CD

Databases & Streaming

Relational, NoSQL, and event-streaming systems — optimized queries, indexing, and high-throughput pipelines.

PostgreSQL

MS SQL Server

MongoDB

MySQL

Kafka Supabase

Architecture & Engineering

Designing systems that stay correct and fast under load — not just code that compiles.

System Design Event-Driven Architecture Distributed Systems

03 Experience

Where I've shipped

4+ years building and owning production systems — from event-driven billing pipelines to full-stack AI platforms serving 50K+ users.

Aug 2025 — Nov 2025

Find Me LLC

New York, USA

● Remote · Internship

Software Developer Intern

Boosted page load 42% (2.4s → 1.4s) and cut payload size 40% by migrating RESTful APIs to GraphQL with field-level resolvers and depth limiting.
Cut manual review effort by 30% by working with operations stakeholders to streamline high-volume review workflows, integrating ML-powered classification into Python backend services.
Shipped customer-requested features across frontend, backend, and AWS deployment pipelines within 2-week sprint cycles.

GraphQL Python ML Integration AWS Docker

Sep 2019 — Jul 2023

iSummation Technologies

India

● On-site · Full-time

Software Developer

Led full-stack platform development (React, Next.js, Spring Boot) serving 50K+ users with 99.8% uptime — achieved through load testing and circuit-breaker patterns.
Enforced OpenAPI contracts across 10+ microservices, cutting production incidents 42% and reducing cross-team integration time from 5 days to 2 days.
Diagnosed and resolved N+1 query patterns and added composite indexes in MS SQL Server, reducing API latency 33% (~900ms → 600ms) across production endpoints.
Mentored 2 junior engineers through structured code reviews and pair-programming, reducing their PR defect rate 40% in Q1.

Java Spring Boot React Next.js MS SQL Microservices

Dec 2018 — Mar 2019

Silverwing Technologies

India

● On-site · Internship

Software Developer Intern

Replaced synchronous billing with a Kafka + Spring Boot event-driven pipeline sustaining 10,000+ events/min with zero message loss and zero billing timeout errors.
Automated payment workflows via Spring Boot REST APIs with schema validation, cutting manual processing effort 40% and eliminating recurring data-entry errors.

Kafka Spring Boot Event-Driven REST APIs

04 Projects

Things I've built

Production-grade AI-powered applications — not tutorials, not clones. Each built end-to-end with real architecture decisions and working deployments.

May 2026

GitHub PR Reviewer

AI-powered code review system that analyzes GitHub pull requests using Claude API. Developers connect through GitHub OAuth, point the service at a PR, and get severity-ranked findings (critical / warning / suggestion) with concrete fixes — surfaced in a results dashboard and posted back as formatted PR comments.

Key Engineering Decisions

GitHub OAuth + API integration — authenticated access to repos, PR metadata, and changed-file diffs
PR state validation — open/closed state checked before analysis so stale PRs are handled correctly
DB-first caching — reviews served from Supabase instantly on repeat requests; bypasses cache when PR is merged or closed
Per-file chunking — each changed file analyzed individually with language detection, preventing context overflow
Structured JSON output — Claude prompted to return strict JSON; parsed into severity-ranked entities stored in PostgreSQL
Persisted history + dashboard — past reviews stored and browsable, with force-refresh to re-run analysis after new commits

Java Spring Boot Next.js TypeScript PostgreSQL Supabase Claude API GitHub OAuth / API Docker

Apr 2026

Intelligent Document Retrieval Platform

Full-stack RAG system for semantic PDF retrieval and context-aware conversational Q&A. Users upload documents and ask questions in natural language — the system retrieves relevant chunks using vector similarity and generates grounded answers using Claude API with real-time token streaming.

Key Engineering Decisions

RAG pipeline — PDFs chunked, embedded, and stored in ChromaDB for semantic vector retrieval
SSE streaming — responses streamed token-by-token to the frontend, reducing perceived latency significantly
Page-level citations — every answer references the specific page it came from
Conversational memory — chat history maintained for multi-turn contextual Q&A

Python FastAPI Next.js ChromaDB Claude API SSE LangChain

Jan 2026

Auth Service (JWT + RBAC)

Authentication service built from the ground up with Spring Boot and Spring Security. Handles registration, login, and role-based access with a refresh-token flow instead of relying on an off-the-shelf auth provider — designed as a foundation that could later be extended into full SSO.

Key Engineering Decisions

JWT access/refresh rotation — short-lived access tokens with rotating refresh tokens, plus refresh-token revocation on logout
Role-based access control — USER/ADMIN roles enforced via Spring Security
BCrypt password hashing and Flyway-managed schema migrations for versioned, repeatable DB setup
Dockerized local setup with GitHub Actions CI

Java Spring Boot Spring Security PostgreSQL JWT RBAC Docker

05 About

The engineer behind the code

I don't just build features.
I own systems.

I'm a Full Stack Engineer with roots in enterprise Java backend development, plus hands-on project experience building production-style AI systems. My career started with event-driven pipelines handling 10K+ events/min at Silverwing, progressed to owning a 50K-user platform at iSummation, and now includes independent projects applying LLM capabilities alongside that backend depth.

What sets me apart is the combination: I understand distributed systems, database performance, and API design from years of backend work — and I've applied that same rigor to my AI projects. I don't just wrap an LLM in a chat interface. I build caching layers, API integration pipelines, structured output parsers, and deployment configurations around them.

I hold an M.S. in Computer Science from Cleveland State University (GPA 3.7), am AWS Certified Cloud Practitioner, and am actively looking for full-stack or AI engineering roles in the US. I'm open to relocating anywhere.

→Systems thinking first — understand the full stack before writing line one
→Metrics-driven — every improvement backed by before/after numbers
→Production mindset — error handling, caching, monitoring, not just happy path
→Fast learner — picked up LangChain, ChromaDB, vector search, and Supabase while building

M.S. Computer Science

Cleveland State University, Ohio

Sep 2023 — May 2025 GPA 3.7/4.0

AWS Certified Cloud Practitioner (CLF-C02)

Amazon Web Services · Aug 2026

🤖

LLM Engineering: Generative AI, RAG, QLoRA & AI Agents

Udemy · Jul 2026

Docker Foundations Professional

Docker, Inc. · Jan 2026

What I build with

Where I've shipped

Things I've built

The engineer behind the code

Let'sconnect.

Let's
connect.