G R HARSHA

Early-stage AI Engineer building production-grade
agentic systems & DevOps copilots.
Open to: AI Engineering Internships · Backend/Platform Roles · DevOps

View Architecture & Code View Résumé

Explore Identity
🚀 Built Gemini DevOps Copilot 🤝 Mentor @ GSSoC 🎓 B.Tech CSE (AI) '29 🤖 Agentic AI Specialist 💻 Python & FastAPI Pro 🌐 Open Source Contributor 🚀 Built Gemini DevOps Copilot 🤝 Mentor @ GSSoC 🎓 B.Tech CSE (AI) '29 🤖 Agentic AI Specialist 💻 Python & FastAPI Pro 🌐 Open Source Contributor
Google Puma PayPal Adobe AMAZON AWS AZURE Google Puma PayPal Adobe

About Me

divider
G R Harsha

Architecture-First AI Engineering

I am an AI Engineer & Agentic Systems Builder focused on moving beyond simple chatbots to autonomous systems that can reason, plan, and execute.

With a background in Computer Science & AI (B.Tech CSE-AI), I specialize in building scalable Microservices, DevOps pipelines, and Multi-Agent Orchestrations. I don't just write code; I design intelligent architectures that solve complex problems.

Explore My Career

Core Competencies

divider

Agentic AI
Systems

Designing autonomous ReAct loops and multi-agent swarms using LangGraph and AutoGen.

DevOps
Copilots

Building AI assistants that generate CI/CD pipelines, K8s manifests, and triage incidents.

AI
Microservices

Deploying scalable model inference endpoints using FastAPI, Docker, and Kubernetes.

Vector
Memory

Implementing RAG pipelines with ChromaDB and FAISS for long-term agent memory.

LLM
Tooling

Integrating function calling (Tools) to connect LLMs with real-world APIs and databases.

Automation
& CLI

Creating developer-first CLI tools that streamline workflows and documentation.

Technical Arsenal

divider
Python & FastAPI Backend & AI
LangChain & AutoGen Orchestration
Docker & K8s Infrastructure
Vector DBs Memory
GitHub Actions CI/CD
Cloud (AWS/GCP) Deployment
Agentic AI Concept
Beyond Chatbots

Autonomous Agentic Systems

I build systems that don't just answer—they act.

My work moves beyond simple RAG pipelines to fully autonomous agentic workflows. By implementing ReAct (Reasoning + Acting) loops and Chain-of-Thought execution, I create agents that can use tools, manage memory, and self-correct.

  • 🔹 Tool Use: Connecting LLMs to APIs, Databases, and DevOps CLI tools.
  • 🔹 Orchestration: Multi-agent swarms using LangGraph and AutoGen.
  • 🔹 Memory: Long-term state management via Vector and Graph databases.

Engineering Approach

divider
System Architecture
Microservices Architecture Pattern for Scalable AI Deployment

Scalability

Designing horizontal scaling strategies using Kubernetes (K8s) and serverless functions.

Security

Implementing RBAC, API Gateways, and secure secret management.

Observability

Full distributed tracing and ELK logging to debug stochastic AI behavior.

Engineered Solutions

divider

Gemini DevOps Copilot

Problem: Manual K8s configs are error-prone.
Outcome: 80% faster config gen.

#FastAPI #Docker #GeminiAPI
Gemini DevOps Copilot

Mentara AI

Problem: Mental health access is expensive.
Outcome: Accessible empathetic AI.

#React #LLM #FullStack
Mentara AI

Code Vortex

Problem: Documentation updates lag code.
Outcome: Automated docs synced to git.

#DevTools #Automation #AI
Code Vortex

Netflix Clone

Pixel-perfect React UI with Firebase backend.

#React #Firebase #UI/UX
Netflix Clone

Object Tracking System

Real-time detection using YOLO/OpenCV.

#Python #OpenCV #ComputerVision
CV Project

Fitness Tracker

Activity analytics dashboard.

#JavaScript #Analytics
Fitness Tracker

Journey

divider

Education

  • B.Tech CSE (Artificial Intelligence) 2025 - 2029 | NIAT & Yenepoya University

    Focusing on Advanced Algorithms, Machine Learning, and Agentic Architectures.

  • Pre-University Presidency PU College

Leadership & Community

  • Mentor GirlScript Summer of Code (GSSoC)

    Guiding open-source contributors in Python and AI projects.

  • Campus Ambassador GUESSS India

Get In Touch

divider

Ready to build the next generation of AI systems? Let's talk.

Harsha AI

close
send