OpsGPT: GitOps-Driven MLOps Inference Engine

OpsGPT is a production-grade inference engine designed to serve Hugging Face LLMs (DistilGPT-2/SmolLM) with high availability and automated delivery.

I built this project to demonstrate a Split-Repository GitOps Strategy, separating application source code from configuration manifests to ensure security, auditability, and clean release cycles.

OpsGPT: GitOps MLOps Pipeline

[ Badges here: CI Status | Python 3.10 | Docker ]

🚀 Overview

OpsGPT is a production-ready LLM inference engine deployed via Kubernetes.

🛠️ Local Development

1. Install dependencies: pip install -r requirements.txt
2. Run server: uvicorn app.main:app --reload

🔮 Future Roadmap

• Add Redis Caching
• Implement Model Quantization

🏗 Architecture

graph LR
    A[Developer] -->|Push Code| B(opsgpt-app Repo)
    B -->|GitHub Action| C{CI Pipeline}
    C -->|Build & Test| D[Docker Registry]
    C -->|Update Tag| E(opsgpt-gitops Repo)
    E -->|Sync| F[ArgoCD]
    F -->|Deploy| G[Kubernetes Cluster]
    G -->|Scale| H[HPA]

Loading