Sethumadhavan N

AI Engineer | Multi-Agent Systems | Thoughtful Builder

Exploring the depths of AI, building intelligent systems, and sharing thoughts along the way

About Me

AI Engineer with internship experience building large language model systems, multi-agent pipelines, and production FastAPI services. Currently developing Doc-it, an open-source CLI for automated codebase documentation. Skilled in agentic frameworks, RAG workflows, and observability platforms.

Core Skills

Languages

PythonSQLJava

Frameworks & Libraries

FastAPILangChainLangGraphREST APIsMicroservicesUvicorn

AI / ML

Scikit-learnPrompt EngineeringRAGAnomaly DetectionMachine Learning

AI Systems

Multi-Agent SystemsObservability AIAI Pipelines

Data Science

PandasNumPyVector methods

Agentic AI

Claude CodeSuperpowersBMAD

Tools

GitGitHubPostmanLinux

Concepts

Model EvaluationFeature EngineeringAI Pipelines

Featured Projects

Doc-it

Open-source CLI that generates structured codebase summaries using a LangGraph orchestration layer with intelligent token budgeting. Delivered a dynamic commit history graph view with pluggable LLM backends and configurable output formats (Markdown, JSON, HTML).

PythonLangChainLangGraphCLI
View on GitHub

AgileCopilot - AI Agile Assistant System

Designing AI assistant for Agile teams (BA, Dev, QA roles). Uses historical sprint data + RAG-style retrieval. Generates acceptance criteria, story points, test cases. Role-based AI workflow design.

PythonLLMRAGAI Pipelines

Agentic War Room POC

Observability platform combining ML anomaly detection with LLM-generated incident narratives for SRE/AIOps teams.

PythonFastAPIMLLLM workflows

LLM Grievance Redressal System

LLM-powered complaint triage system that classifies citizen grievances and automatically routes each case to the appropriate government authority.

PythonLLMClassification

Experience & Education

Work Experience

AI Engineer Intern

Primesoft Inc. · Chennai

March 2026 – Present

  • Engineered a RAG system using vector embeddings and semantic similarity search, achieving sub-2s query resolution for a voice-based AI agent
  • Developed and integrated REST APIs using Node.js and FastAPI; leveraged AI-assisted tooling (Claude Code) to accelerate development velocity

AI Research Intern

BluBridge AI Research Lab · Chennai

Aug 2025

  • Studying AI systems stack including compute execution, kernels, compiler concepts, and parallelism
  • Working with systems team on model execution efficiency and infrastructure fundamentals
  • Reading and implementing low-level architecture concepts behind deep learning frameworks
  • Researching LLM architecture components, tokenization, activation functions, and training pipeline math
  • Contributing to internal AI infra understanding and optimization discussions

AI Engineer Intern

AutonomOps · Hyderabad

March 2025 – June 2025

  • Built 4+ FastAPI microservices for a production observability platform, ingesting logs and metrics across distributed systems in real time
  • Architected a multi-agent anomaly detection workflow using z-score and TF-IDF statistical models, reducing false-positive alert rate by over 40%
  • Integrated LLM workflows to generate structured JSON incident summaries for direct dashboard rendering, cutting mean triage time by approximately 30%

Education

B.Tech — Artificial Intelligence & Data Science

SRM Valliammai Engineering College · Chennai

2022 – 2026

CGPA: 8.83

CBSE Class 12

Sri.R.M.Jain Vidhyasharam · Thiruvallur

2022

93.4%

CBSE Class 10

Sri.R.M.Jain Vidhyasharam · Thiruvallur

2020

94.6%

Certifications

The Joy of Computing in Python

NPTEL

81%

Data Mining

NPTEL

79%

Data Structure and Algorithms using Java

NPTEL

72%

Database Management and Systems

NPTEL

69%

Achievements

Winner of Makeathon 2025

VIT Chennai

February 2025

Developed AI-driven code mentorship Python IDE with Electron.js integrated with Gemini API. Secured internship with Innova Solutions.

2nd Place in Data Analytics Hackathon

IIITDM, Kancheepuram

Built ML models and Power BI dashboards for analyzing financial datasets and uncovering hidden trends.

Runner-Up in Hackathon

Amrita Vishwa Vidyapeetham

March 2024

Developed Automated Desoldering Machine for E-waste management using Arduino.

You've reached the bottom

Let's Connect

Interested in collaborating or just want to say hi? Reach out!

Built with Next.js, Tailwind CSS, and Framer Motion
Created using Antigravity
© 2026 Sethumadhavan N