Loading Portfolio…
MANOJ
MANOJDEVARAJULU
— Chennai, India

Building intelligent systems.

I engineer AI infrastructure, developer tools, and intelligent workflows. Systems-first thinker. Open to Work.

Open to Work — Available for new projects & roles
Python TypeScript JavaScript FastAPI React Next.js Node.js LLMs RAG / Agentic LangGraph Azure PostgreSQL MongoDB Git GitHub Docker Vercel Supabase Python TypeScript JavaScript FastAPI React Next.js Node.js LLMs RAG / Agentic LangGraph Azure PostgreSQL MongoDB Git GitHub Docker Vercel Supabase 

I think in systems, not features.

Hi, I'm Manoj D — an AI/ML Engineer with 1.5+ years building production AI applications, LLM-powered automation, agentic workflows, and RAG systems. I think in systems rather than isolated features, always focused on reducing complexity and delivering measurable impact.

At Infinite Computer Solutions, I architected Zen AI — a production OCR + LLM platform for healthcare document processing that cut processing time by ~90% and saved ~400 hours/month. I design HIPAA-compliant PHI pipelines, Azure-based cloud architectures, and structured prompt-engineering workflows.

On the side, I'm building GravLens-MMA, an AI pipeline that ranks galaxies for gravitational-wave telescope follow-up — and documenting the process on Medium. I also built Sentinel (RAG + dependency graph developer copilot) and an agentic chess engine with Minimax AI.

PythonFastAPILangGraphRAGGPT-4oAzurePostgreSQLAstropy

“Technology should reduce complexity, not add to it. Every system I build starts from that premise.”

— Engineering philosophy
1.5+
Years in production AI systems
~90%
Processing time reduced at Zen AI
Systems still left to build

Projects that matter.

Each project started from a real frustration, a genuine problem, or a system I wished existed.

FeaturedIn Progress01Astro AI Research

GravLens-MMA — AI Gravitational-Wave Follow-up Pipeline

Building an AI-powered ranking pipeline for gravitational-wave telescope follow-up using Python, PostgreSQL, Astropy, HEALPix, and the GLADE+ catalog. Focuses on scalable data ingestion, feature engineering, and graph-based machine learning ranked by Hit Rate@K and MRR.

PythonAstropyHEALPixPostgreSQLGraph ML
Helps telescopes decide where to look — ranking the galaxy most likely to host a gravitational wave source
02Healthcare AI

Zen AI — Healthcare Automation Platform

A production OCR + LLM platform that automates healthcare referral document processing. Reduces processing time by ~90% and saves ~400 hours/month. Built with HIPAA-compliant PHI classification, Azure blob storage, and structured prompt-engineering pipelines.

GPT-4oTesseract OCRFastAPIAzure
~90% reduction in processing time · ~400 hrs/month saved
[ Open Direct Link ↗ ]
03Dev Infrastructure

Sentinel — Developer Intelligence Copilot

A developer copilot that combines dependency graph analysis with Retrieval-Augmented Generation (RAG) to enable natural-language code search. Maps architectures, visualizes module dependencies, and helps developers navigate large codebases conversationally.

Next.jsFastAPIRAGDependency Graphs
Improves developer cognition and codebase navigability
[ Open Direct Link ↗ ]
In Progress04Game Ecosystem

En Passant — Territory Chess Ecosystem

A territory-based chess ecosystem combining competitive chess strategy, geolocation, real-time multiplayer interactions, and persistent world mechanics. Python chess engine with full rule enforcement and Minimax AI using alpha-beta pruning.

PythonGame SystemsMinimax AIMultiplayer
Competitive chess strategy mapped to real-world geolocation
In Progress

Where I've worked.

Exposure to regulated healthcare environments taught me the importance of reliability, security, and production-grade AI deployment — things you can't learn from tutorials.

Associate Software Engineer

Oct 2024 — Present
Infinite Computer Solutions

Enterprise healthcare AI systems — architected Zen AI, a production OCR + LLM platform that automates referral document processing.

~90%Time Saved
~400 hrsSaved / month
  • Built OCR + LLM pipelines delivering precise, HIPAA-compliant document processing
  • Implemented secure PHI classification, ingestion and structured JSON delivery
  • Azure Blob Storage, Azure Data Factory migration, and custom Selenium test automation

Full-Stack Developer Intern

Jul 2024 — Sep 2024
Inblue Infotech Pvt Ltd

Built full-stack product features across the React + Node.js stack. Contributed to REST API design, MongoDB schemas, code reviews, and deployment workflows.

  • Full-stack features with React, Node.js, REST APIs, and MongoDB
  • Participated in code reviews and production deployment workflows

What I'm doing right now.

A live snapshot. Updated whenever something meaningful shifts.

Building
GravLens-MMA
AI ranking pipeline for gravitational-wave telescope follow-up — deciding where in the sky telescopes should look
Thinking About
LLM Reliability Patterns
How multi-agent systems fail in production and what actually makes agentic workflows robust
Writing On
Medium
Documenting the build process: AI for astrophysics, developer tools, and systems thinking in public

Tools I actually use.

No buzzword bingo. Only tools I've shipped real things with.

Core: Python + FastAPI for AI backends. Next.js + React for product layers. LangGraph for agentic workflows. Azure + Docker for cloud infrastructure.

Python
TypeScript
JavaScript
FastAPI
React
Next.js
Node.js
LLMs
RAG / Agentic
LangGraph
Azure
PostgreSQL
MongoDB
Git
GitHub
Docker
Vercel
Supabase

Where I learned to build.

A background in AI & Data Science gave me the theoretical foundations to build things that actually work in production — not just demo well.

B.Tech — Artificial Intelligence & Data Science

2020 — 2024
Velammal Engineering College, Chennai
CGPA: 8.88 / 10Anna University Affiliated

Specialized in machine learning, neural networks, NLP, and data engineering. Built the foundation for everything from OCR pipelines to agentic AI systems.


Let's build
something.

Open to interesting conversations — AI systems, startup ideas, developer tooling, collaborations, or just a good technical discussion.

builds@manojdevarajulu.cc
Open to Work — Available

Based in Chennai, India · Usually responds within 24 hours