Sharath S · Full-Stack AI Engineer

Full-Stack AI Engineer

Building agentic AI systems, end to end.

I build multi-agent systems with LangChain & LangGraph, production RAG on Weaviate, and ship AI features in NestJS/Angular for European enterprise clients.

Thiruvananthapuram, Kerala, IndiaOpen to senior roles

ai & retrieval

LangChain
LangGraph
RAG pipelines
Prompt engineering
Vector embeddings
Multi-agent orchestration
Agentic AI
Semantic search
Model Context Protocol (MCP)
SSE streaming
OpenAI API
Anthropic Claude API
Google Gemini API
Google ADK

backend

Node.js
NestJS
Express.js
FastAPI
REST APIs
GraphQL
JWT auth
Socket.IO
RabbitMQ

frontend

React
Angular
Next.js
Vue.js
Tailwind CSS
SASS / SCSS

databases

PostgreSQL
MongoDB
Redis
MySQL
Weaviate
Elasticsearch

cloud & devops

Docker
Kubernetes
AWS
Azure
Terraform
Nginx
GitHub Actions (CI/CD)

testing & observability

Jest
Unit testing
E2E testing
Langfuse

tools & methods

Git
Agile / Scrum
Microservices
Event-driven architecture
Claude Code
GitHub Copilot
Cursor

languages

JavaScript
TypeScript
Python
HTML
CSS

Dec 2024 → June 20261 yr 7 mosSurat, Gujarat, India · On-site

Full-Stack Developer

Dignizant Technologies LLP (Client: elunic GmbH)

Build full-stack features for ShopfloorGPT, an AI-powered industrial knowledge platform for European enterprise clients, on an Angular/Vue.js/NestJS stack with multi-tenant auth and role-based access control.

Built full-stack features for ShopfloorGPT (industrial knowledge platform) using Angular, Vue.js, NestJS, and TypeScript, with multi-tenant authentication and role-based access control across tenants.
Engineered production RAG pipelines — document ingestion, hierarchical chunking, vector embeddings, and Weaviate semantic search — for enterprise knowledge retrieval across multiple tenants.
Built LangGraph multi-agent workflows with a decorator-based plugin system for tool discovery and sub-agent delegation, applying LLM-as-router orchestration patterns.

AngularVue.jsNestJSTypeScriptPayloadCMSPostgreSQLMongoDBWeaviateLangChainLangGraphDockerAzureTraefikNX MonorepoOpenAPIGitLab CILangfuse

Nov 2022 → Nov 20242 yrsThiruvananthapuram, Kerala, India · Remote

MERN Developer

Freelance (Self-Employed)

Built and shipped e-commerce platforms and real-time apps on the MERN stack for clients, with payments, real-time features, and cloud deployment.

Built and shipped e-commerce platforms with the MERN stack (MongoDB, Express, React, Node.js), integrating Stripe and PayPal payment flows plus real-time chat and notifications via WebSockets and Socket.IO.
Designed event-driven microservices with Docker, Kubernetes, and RabbitMQ, and deployed to AWS (EC2, S3, RDS) with CI/CD via Jenkins.
Optimized performance through Redis caching and Bull-based asynchronous job processing.

ReactNode.jsExpress.jsMongoDBRedisBullSocket.IOTypeScriptStripePayPalAWSDockerKubernetesRabbitMQJenkins

#01

Multi-Agent Research Assistant — LangGraph Orchestration

personal·2026

A multi-agent research system built on LangGraph that turns a single question into a reviewed markdown report. Four cooperating agents — Supervisor, Researcher, Writer, Reviewer — are nodes in a stateful StateGraph sharing a typed Pydantic v2 state object merged via a reducer. A supervisor/orchestrator-worker pattern routes dynamically through conditional edges (the next agent is chosen at runtime from shared state, not a hard-coded pipeline). A Writer–Reviewer refinement loop iterates with a revision cap (max 3) for guaranteed termination, and live web research runs through a keyless DuckDuckGo tool plus an LLM-backed summarizer. A provider-agnostic LLM layer (Groq Llama 3.3 70B, Google Gemini, OpenAI, Anthropic Claude) is switchable via one env var, with node-by-node streaming output and a multi-stage Docker build.

pythonlanggraphlangchainmulti-agentagentic aipydanticdockerrag

↗ details github

#02

PDF Chat Assistant — RAG with OCR & Visual Citations

personal·2026

A production-style RAG chatbot that answers questions grounded in user PDFs — including scanned/image-only PDFs handled per page via Tesseract OCR (triggered only when a page has <50 native chars, to avoid needless OCR cost). Its signature feature is visual, verifiable citations: every retrieved chunk maps back to bounding boxes overlaid on the rendered source page, computed at processing time (pages rendered at 2× zoom) and carried as chunk metadata, so answers point to exactly where they came from. Semantic retrieval uses Sentence-Transformers (all-MiniLM-L6-v2) embeddings in a FAISS store with a scikit-learn TF-IDF fallback, multi-turn memory, and resilient generation (fallback, retries, 2s rate-limit throttling for free-tier Gemma ~30 RPM).

pythonraglangchaingoogle geminifaisssentence-transformerstesseract ocrstreamlitdocker

↗ details github

#03

Resume-JD Tailor — Multi-Agent Resume Tailoring Service

personal·2026

A multi-agent service that scores a resume against a job description skill-by-skill (0–100 fit, covered/partial/missing with evidence) and rewrites bullets to fit the role — flagging genuinely missing skills as honest gaps instead of fabricating experience. A four-stage pipeline on Google ADK with Gemini uses sequential + parallel orchestration: a ParallelAgent parses JD and resume concurrently (gemini-2.5-flash-lite), a match analyzer scores fit, and a tailor agent rewrites (gemini-2.5-flash) — passing data through ADK session state with Pydantic schema validation at every handoff. Model routing uses cheap flash-lite for 3 of 4 stages and the stronger model only for the rewrite. Input handling is hardened with PDF magic-byte validation, a 10 MB cap, encryption/scanned-PDF rejection (HTTP 422 before any LLM call), and a zero-fabrication evaluation tripwire (eval.py over labeled cases) enforced as an acceptance gate.

pythonfastapigoogle adkgoogle geminimulti-agentpydanticdocker

↗ details github

#04

WorkHub — Microservice-Based Job Application Platform

personal·2024

A scalable freelance/job-application platform built on a microservices, event-driven architecture with Node.js, TypeScript, Docker, and Kubernetes. Comprises 6 core microservices — API Gateway, Authentication, Gig CRUD, Order/Payment (Stripe), Notification, and real-time Chat — with Elasticsearch-powered search and filtering across gigs.

node.jstypescriptmicroservicesrabbitmqelasticsearchstripedockerkubernetes

↗ details github

#05

Social Media Platform — Real-Time Messaging & Content System

personal·2024

A full-featured social media backend on Node.js + TypeScript + Express + MongoDB, with user authentication, post management, real-time private messaging via Socket.io, and background job processing with Bull and Redis on AWS.

node.jstypescriptmongodbredisbullsocket.ioawsjest

↗ details github

#06

ElectroMart — Full-Stack E-Commerce Platform

personal·2024

A feature-complete e-commerce SPA built with React + Redux Toolkit (Vite) and a Node.js/Express/MongoDB API. Includes a paginated, searchable product catalog with a carousel, shopping cart, product reviews/ratings, user profiles with order history, a full checkout flow (shipping → payment), order tracking, and an admin area for managing products, users, and orders. Auth is JWT + bcrypt over httpOnly cookies with role-based admin vs. customer access; payments via PayPal checkout (incl. card payments through PayPal).

reactreduxnode.jsexpressmongodbjwtpaypalbootstrap

↗ details github

Aug 2014 → Jul 2018Thiruvananthapuram, Kerala, India

B.Tech in Computer Science and Engineering

Rajadhani Institute of Engineering and Technology

You Can Only Test What You Kept Deterministic: pytest for a LangGraph Agent · 7/9/2026
A LangGraph agent test suite: two seconds, no API key, zero LLM calls. Every invariant moved out of the prompt becomes a function you can pytest.
The 429 Is a UX Problem: Shipping an LLM Demo on Free-Tier Keys · 7/9/2026
Rate limits on free-tier keys aren't an edge case — they're the weather. Classifying two kinds of 429, honest retry times, and a provider fallback.
Zero Fabrication by Construction: A Multi-Agent Resume Tailor on Google ADK · 7/7/2026
An agent pipeline that rewrites your resume for a job description — and is structurally incapable of inventing experience you don't have. Honesty enforced by an eval tripwire that must read zero, not by a hopeful prompt.
Trustworthy RAG: Bounding-Box Citations and Conditional OCR · 7/6/2026
A RAG chatbot you can verify at a glance — every answer highlights the exact region of the page it came from, scanned PDFs get OCR only when a page needs it, and every layer has a fallback.
Cite or Refuse: Debugging a Fluent Hallucination in a LangGraph Research Agent · 7/5/2026
My multi-agent research assistant produced a polished report full of fabricated APIs — and it looked MORE trustworthy the longer it ran. The fix wasn't a better prompt: evidence-carrying state, a deterministic validation gate, and a cite-or-refuse contract.

Ask me anything about this portfolio — grounded in the projects, experience, and posts below. Type chat <your question> in the terminal above (e.g. chat what has he built with rag?) and the answer streams back with citations.

emailsharath.surendran.m@gmail.com

phone+91 70251 27271

githubgithub.com/ki11e6

linkedinlinkedin.com/in/sharath-s-001590137

locationThiruvananthapuram, Kerala, India

statusOpen to senior roles

languagesMalayalam (Native) · English (Professional) · Tamil (Elementary)

Introduction

$

Skills

ai & retrieval

backend

frontend

databases

cloud & devops

testing & observability

tools & methods

languages

Experience

Full-Stack Developer

MERN Developer

Projects

Multi-Agent Research Assistant — LangGraph Orchestration

PDF Chat Assistant — RAG with OCR & Visual Citations

Resume-JD Tailor — Multi-Agent Resume Tailoring Service

WorkHub — Microservice-Based Job Application Platform

Social Media Platform — Real-Time Messaging & Content System

ElectroMart — Full-Stack E-Commerce Platform

Education

B.Tech in Computer Science and Engineering

Blog

AI assistant

Contact