PodSage-AI

🧠 PodSage AI

AI-Powered Kubernetes Observability & Infrastructure Intelligence

Real-time telemetry β€’ AI-driven anomaly detection β€’ Operational intelligence

🏒 About PodSage AI

PodSage AI is the organization behind this repository. The PodSage-AI project is a flagship open-source platform built by PodSage AI to deliver intelligent Kubernetes observability, AI-assisted anomaly detection, and infrastructure intelligence.

πŸ“– Overview

PodSage-AI is a flagship project built by the PodSage AI organization. It is an intelligent Kubernetes observability platform that monitors, analyzes, and correlates real-time infrastructure behavior using AI-powered operational insights.

The project includes a React/Vite frontend dashboard for interactive metric visualization, dependency maps, and AI insights.

Built for the ABB Accelerator 2026 challenge, PodSage-AI combines Kubernetes telemetry, Prometheus metrics, anomaly detection, dependency analysis, and infrastructure intelligence into a unified monitoring ecosystem.

The mission is simple:

Transform raw Kubernetes metrics into actionable operational intelligence.

❓ Why PodSage AI?

Traditional observability platforms expose metrics.

PodSage AI focuses on transforming telemetry into actionable operational intelligence using AI-assisted infrastructure analysis.

Instead of only showing dashboards, PodSage AI helps explain:

✨ Core Features

πŸ—οΈ System Architecture

flowchart LR

    A["Applications / Microservices"]

    B["Data Collection Layer
    β€’ Prometheus
    β€’ Node Exporter
    β€’ kube-state-metrics
    β€’ cAdvisor"]

    C["AI Intelligence Layer
    β€’ CPU Analysis Engine
    β€’ Memory Analysis Engine
    β€’ Dependency Mapper
    β€’ Correlation Engine"]

    D["Infrastructure Intelligence Layer
    β€’ Prometheus
    β€’ SQLite
    β€’ Loki
    β€’ ML Models"]

    E["Dashboard & Visualization Layer
    β€’ React / Vite
    β€’ Recharts
    β€’ React Flow
    β€’ WebSockets"]

    A --> B
    B --> C
    C --> D
    D --> E

βš™οΈ Tech Stack

Backend

Monitoring & Metrics

Infrastructure

AI & Analysis

Frontend

πŸ“ Project Structure

PodSage-AI/
β”œβ”€β”€ backend/
β”‚   β”œβ”€β”€ app/
β”‚   β”‚   β”œβ”€β”€ api/
β”‚   β”‚   β”œβ”€β”€ database/
β”‚   β”‚   β”œβ”€β”€ models/
β”‚   β”‚   β”œβ”€β”€ services/
β”‚   β”‚   β”œβ”€β”€ websocket/
β”‚   β”‚   └── main.py
β”‚   β”‚
β”‚   β”œβ”€β”€ Dockerfile
β”‚   β”œβ”€β”€ docker-compose.yml
β”‚   β”œβ”€β”€ prometheus.yml
β”‚   β”œβ”€β”€ requirements.txt
β”‚   └── podsage.db
β”œβ”€β”€ frontend/
β”‚   β”œβ”€β”€ index.html
β”‚   β”œβ”€β”€ package.json
β”‚   β”œβ”€β”€ README.md
β”‚   β”œβ”€β”€ vite.config.js
β”‚   β”œβ”€β”€ src/
β”‚   β”‚   β”œβ”€β”€ App.jsx
β”‚   β”‚   β”œβ”€β”€ main.jsx
β”‚   β”‚   β”œβ”€β”€ api/
β”‚   β”‚   β”‚   └── client.js
β”‚   β”‚   β”œβ”€β”€ components/
β”‚   β”‚   β”‚   β”œβ”€β”€ AIInsights.jsx
β”‚   β”‚   β”‚   β”œβ”€β”€ AnomalyTable.jsx
β”‚   β”‚   β”‚   β”œβ”€β”€ ClusterSummary.jsx
β”‚   β”‚   β”‚   β”œβ”€β”€ DependencyGraph.jsx
β”‚   β”‚   β”‚   β”œβ”€β”€ Header.jsx
β”‚   β”‚   β”‚   β”œβ”€β”€ JsonPreview.jsx
β”‚   β”‚   β”‚   β”œβ”€β”€ MetricCard.jsx
β”‚   β”‚   β”‚   └── SeriesChart.jsx
β”‚   β”‚   β”œβ”€β”€ styles/
β”‚   β”‚   β”‚   └── global.css
β”‚   β”‚   └── utils/
β”‚   β”‚       └── metrics.js
β”œβ”€β”€ README.md
β”œβ”€β”€ LICENSE
└── .gitignore

πŸš€ Getting Started

Prerequisites

Before starting, ensure you have:

πŸ“¦ Installation

1. Clone Repository

git clone https://github.com/PodSageAI/PodSage-AI.git
cd PodSage-AI/backend

2. Create Virtual Environment

python -m venv venv

Linux / macOS

source venv/bin/activate

Windows

venv\Scripts\activate

3. Install Dependencies

pip install -r requirements.txt

4. Frontend Setup

cd ../frontend
npm install

▢️ Running the Backend

Local Development

uvicorn app.main:app --reload

Backend URL:

http://localhost:8000

Swagger Documentation:

http://localhost:8000/docs

ReDoc Documentation:

http://localhost:8000/redoc

▢️ Running the Frontend

Local Development

cd frontend
npm run dev

Frontend URL:

http://localhost:5173

πŸŽ₯ Demo

Live Metrics API

curl http://localhost:8000/metrics/cpu

Open Swagger UI

http://localhost:8000/docs

🐳 Docker Usage

Start Services

docker compose up --build

Stop Services

docker compose down

Run in Detached Mode

docker compose up -d

☸️ Kubernetes Deployment

Apply Kubernetes Resources

kubectl apply -f k8s/

Verify Pods

kubectl get pods

Port Forward Backend

kubectl port-forward svc/podsage-ai 8000:8000

πŸ“‘ API Endpoints

Health Endpoints

Endpoint Description
/ Root status
/health Health check

Metrics Endpoints

Endpoint Description
/metrics/cpu CPU metrics
/metrics/memory Memory metrics
/metrics/restarts Restart metrics

AI & Intelligence Endpoints

Endpoint Description
/anomalies Detected anomalies
/insights AI-generated insights
/dependencies Dependency mapping

πŸ“˜ Example API Responses

CPU Metrics

{
  "status": "success",
  "data": {
    "resultType": "vector",
    "result": [
      {
        "metric": {},
        "value": [
          1778683850.411,
          "0.2482235237555631"
        ]
      }
    ]
  }
}

Anomaly Detection

[
  {
    "type": "High CPU Usage",
    "pod": "node-exporter:9100",
    "value": 24.82,
    "unit": "%"
  }
]

AI Insights

[
  {
    "pod": "node-exporter:9100",
    "insight": "Pod node-exporter:9100 is consuming unusually high CPU resources.",
    "recommendation": "Consider scaling replicas or optimizing workload."
  }
]

🧠 AI Capabilities

Current AI functionality includes:

Default Thresholds

CPU_THRESHOLD = 0.2
MEMORY_THRESHOLD = 500000000
RESTART_THRESHOLD = 5

πŸ›‘οΈ Fault-Tolerant Monitoring

PodSage AI automatically falls back to node-level metrics when container-level Kubernetes metrics are unavailable.

This ensures monitoring continuity even in partially configured environments.

Example fallback query:

1 - avg(rate(node_cpu_seconds_total{mode="idle"}[1m]))

πŸ“Š Observability Workflow

  1. Kubernetes metrics are scraped via Prometheus
  2. Metrics are processed by intelligence services
  3. Infrastructure anomalies are detected
  4. Correlation engine generates operational insights
  5. Real-time updates stream through WebSockets
  6. Dashboards visualize cluster intelligence

βœ… Current Capabilities

πŸ§ͺ Example Use Cases

πŸ›£οΈ Roadmap

πŸ† ABB Accelerator 2026

PodSage AI was developed as part of the ABB Accelerator 2026 innovation challenge focused on:

🀝 Contributing

Contributions are welcome.

Steps to Contribute

1. Fork the Repository

2. Create a Feature Branch

git checkout -b feature/my-feature

3. Commit Changes

git commit -m "Add new feature"

4. Push to Branch

git push origin feature/my-feature

5. Open a Pull Request

πŸ‘₯ Maintainers

οΏ½ Release Notes

PodSage AI v0.1.4-alpha

Improvements

Frontend Enhancements

Backend Enhancements

Infrastructure

πŸ“„ License

MIT License Β© 2026 PodSage AI

πŸ“Œ Project Status

Version: v0.1.4-alpha
Status: Active Development

🌟 Vision

PodSage AI aims to evolve into a next-generation autonomous infrastructure intelligence platform capable of understanding, predicting, and optimizing Kubernetes environments in real time.

Future versions aim to transition from observability into fully autonomous operational intelligence.

Built with ❀️ for cloud-native infrastructure intelligence

This site is open source. Improve this page.