# 🌟 OpenFinOps
**Open Source FinOps Platform for AI/ML Cost Observability and Optimization**
[](https://opensource.org/licenses/Apache-2.0)
[](https://www.python.org/downloads/)
[](https://pypi.org/project/openfinops/)
[](https://pypi.org/project/openfinops/)
[](https://pepy.tech/project/openfinops)
[](CONTRIBUTING.md)
## 🎯 Overview
OpenFinOps is a comprehensive, open-source platform for tracking, analyzing, and optimizing costs across AI/ML infrastructure and operations. It provides real-time visibility into:
- **LLM Training Costs** - Track GPU utilization, training jobs, and compute expenses
- **RAG Pipeline Monitoring** - Monitor vector databases, embeddings, and retrieval costs
- **Multi-Cloud Cost Tracking** - Unified view across AWS, Azure, GCP
- **AI API Usage** - Track OpenAI, Anthropic, and other LLM API costs
- **Executive Dashboards** - Role-based views for CFO, COO, Infrastructure Leaders
- **Cost Attribution** - Per-model, per-team, per-project cost breakdown
- **AI-Powered Optimization** - Intelligent recommendations for cost savings
## ✨ Key Features
### 📊 Comprehensive Observability
- **Real-time Monitoring**: Live metrics for all AI/ML operations
- **Multi-Cloud Support**: AWS, Azure, GCP telemetry integration
- **LLM Training Tracking**: GPU utilization, loss curves, cost per epoch
- **RAG Analytics**: Document processing, embedding generation, retrieval accuracy
- **API Cost Tracking**: OpenAI, Anthropic, and custom API endpoints
### 💰 FinOps Intelligence
- **Cost Observatory**: Centralized cost tracking and analysis
- **Cost Attribution**: Per-model, per-project, per-team breakdowns
- **Budget Management**: Set budgets, track spending, get alerts
- **Optimization Recommendations**: AI-powered cost-saving suggestions
- **Trend Analysis**: Historical cost patterns and forecasting
### 📈 Professional Visualizations
- **VizlyChart**: Built-in visualization library for charts and graphs
- **Executive Dashboards**: Role-based dashboards (CFO, COO, Infrastructure)
- **Real-time Charts**: Live updating metrics and KPIs
- **Custom Reports**: Generate and export custom cost reports
### 🔐 Enterprise-Grade Security
- **Role-Based Access Control (RBAC)**: Fine-grained permissions
- **IAM Integration**: Identity and access management system
- **Audit Logging**: Complete audit trail for all operations
- **Secure API**: Authentication and authorization built-in
### 🚀 Easy Deployment
- **Docker Support**: Containerized deployment ready
- **Kubernetes Ready**: Helm charts for K8s deployment
- **Cloud-Native**: Deploy on any cloud provider
- **On-Premise**: Run in your own datacenter
## 🏗️ Architecture
```
openfinops/
├── observability/ # Core observability platform
│ ├── observability_hub.py # Central monitoring hub
│ ├── llm_observability.py # LLM training & RAG monitoring
│ ├── finops_dashboards.py # FinOps dashboards
│ ├── cost_observatory.py # Cost tracking
│ ├── cost_reporting.py # Cost reporting
│ ├── ai_recommendations.py # AI-powered optimization
│ └── alerting_engine.py # Intelligent alerting
│
├── vizlychart/ # Visualization library
│ ├── charts/ # Chart implementations
│ ├── rendering/ # Rendering engine
│ └── core/ # Core utilities
│
├── dashboard/ # Role-based dashboards
│ ├── cfo_dashboard.py # CFO financial view
│ ├── coo_dashboard.py # COO operations view
│ ├── infrastructure_leader_dashboard.py
│ └── iam_system.py # Access control
│
└── agents/ # Cloud telemetry agents
├── aws_telemetry_agent.py
├── azure_telemetry_agent.py
├── gcp_telemetry_agent.py
└── generic_telemetry_agent.py
```
## 🚀 Quick Start
### Installation
```bash
# Install from source
git clone https://github.com/openfinops/openfinops.git
cd openfinops
pip install -e .
# Or with all features
pip install -e ".[all]"
# Or specific cloud providers
pip install -e ".[aws,azure,gcp]"
```
### Basic Usage
```python
from openfinops.observability import ObservabilityHub
from openfinops.observability import LLMObservabilityHub
# Initialize observability
hub = ObservabilityHub()
llm_hub = LLMObservabilityHub()
# Register training cluster
llm_hub.register_training_cluster(
cluster_name="gpu-cluster-1",
nodes=["node-1", "node-2"]
)
# Track training metrics
llm_hub.track_training_metrics(
model_id="gpt-custom",
epoch=1,
step=100,
loss=0.5,
gpu_memory_usage=8000
)
# Get cost summary
cost_summary = hub.get_cost_summary()
print(f"Total spend: ${cost_summary['total']}")
```
### Running the Dashboard
```bash
# Start the web dashboard
openfinops-dashboard
# Or with custom port
openfinops-dashboard --port 8080
# Access at http://localhost:8080
```
## 📚 Documentation
- **[Installation Guide](docs/installation.md)** - Detailed installation instructions
- **[Quick Start](docs/quickstart.md)** - Get started in 5 minutes
- **[API Reference](docs/api-reference.md)** - Complete API documentation
- **[Architecture](docs/architecture.md)** - System architecture overview
- **[Deployment Guide](docs/deployment.md)** - Production deployment guide
- **[Contributing](CONTRIBUTING.md)** - How to contribute
## 🎯 Use Cases
### For Engineering Teams
- Monitor LLM training job costs in real-time
- Track GPU utilization and optimize resource allocation
- Debug cost anomalies and inefficiencies
- Set up alerts for budget overruns
### For Finance Teams
- Get complete visibility into AI/ML spending
- Track cost attribution by team, project, or model
- Generate reports for stakeholders
- Forecast future AI infrastructure costs
### For Leadership
- Executive dashboards with key metrics
- ROI analysis for AI initiatives
- Budget vs. actual spending tracking
- Strategic cost optimization recommendations
## 🌐 Multi-Cloud Support
OpenFinOps supports telemetry from:
- **AWS**: CloudWatch metrics, Cost Explorer, EC2, SageMaker
- **Azure**: Azure Monitor, Cost Management, Azure ML
- **GCP**: Cloud Monitoring, Cloud Billing, Vertex AI
- **AI Platforms**: OpenAI, Anthropic, Hugging Face, Custom APIs
## 🤝 Contributing
We welcome contributions! Please see our [Contributing Guide](CONTRIBUTING.md) for details.
### Development Setup
```bash
# Clone the repository
git clone https://github.com/openfinops/openfinops.git
cd openfinops
# Create virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install development dependencies
pip install -e ".[dev]"
# Run tests
pytest
# Format code
black src/ tests/
ruff check src/ tests/
```
## 📊 Roadmap
- [ ] Kubernetes operator for auto-scaling based on cost
- [ ] Integration with Prometheus and Grafana
- [ ] Support for more AI platforms (Cohere, AI21, etc.)
- [ ] Advanced anomaly detection with ML models
- [ ] Mobile app for cost monitoring
- [ ] Slack/Teams integration for alerts
- [ ] Custom webhook support
- [ ] Cost allocation tagging system
## 📄 License
OpenFinOps is licensed under the Apache License 2.0. See [LICENSE](LICENSE) for details.
## 🙏 Acknowledgments
OpenFinOps is built with inspiration from the FinOps Foundation principles and best practices from the cloud cost optimization community.
## 📝 Citing OpenFinOps
If you use OpenFinOps in your research, please cite it using the following references:
### BibTeX
```bibtex
@software{openfinops2024,
title = {{OpenFinOps}: Open Source FinOps Platform for AI/ML Cost Observability and Optimization},
author = {Durai and {OpenFinOps Contributors}},
year = {2024},
month = {10},
version = {0.1.0},
url = {https://github.com/rdmurugan/OpenFinOps},
license = {Apache-2.0},
keywords = {finops, cost-optimization, observability, ai-ml, cloud-cost, llm-monitoring}
}
## 📞 Support
- **GitHub Issues**: [Report bugs or request features](https://github.com/openfinops/openfinops/issues)
- **Discussions**: [Community discussions](https://github.com/openfinops/openfinops/discussions)
- **Email**: durai@infinidatum.net
## ⭐ Star Us!
If you find OpenFinOps useful, please consider giving us a star on GitHub! It helps the project grow and reach more users.
---
**Made with ❤️ by the OpenFinOps community**
Raw data
{
"_id": null,
"home_page": null,
"name": "openfinops",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": "OpenFinOps Contributors <durai@infinidatum.net>",
"keywords": "finops, cost-optimization, observability, ai-ml, cloud-cost, aws, azure, gcp, openai, anthropic, llm-monitoring, rag-monitoring, cost-attribution, visualization, dashboard, telemetry, multi-cloud, intelligent-recommendations, hardware-optimization, gpu-optimization, scaling-recommendations, auto-scaling, llm-powered, claude",
"author": null,
"author_email": "OpenFinOps Contributors <durai@infinidatum.net>",
"download_url": "https://files.pythonhosted.org/packages/8c/d9/61f5e8f61d74327079a9a5bd4e713a0ece45c13a4982e66149a862317244/openfinops-0.2.1.tar.gz",
"platform": null,
"description": "# \ud83c\udf1f OpenFinOps\n\n**Open Source FinOps Platform for AI/ML Cost Observability and Optimization**\n\n[](https://opensource.org/licenses/Apache-2.0)\n[](https://www.python.org/downloads/)\n[](https://pypi.org/project/openfinops/)\n[](https://pypi.org/project/openfinops/)\n[](https://pepy.tech/project/openfinops)\n[](CONTRIBUTING.md)\n\n## \ud83c\udfaf Overview\n\nOpenFinOps is a comprehensive, open-source platform for tracking, analyzing, and optimizing costs across AI/ML infrastructure and operations. It provides real-time visibility into:\n\n- **LLM Training Costs** - Track GPU utilization, training jobs, and compute expenses\n- **RAG Pipeline Monitoring** - Monitor vector databases, embeddings, and retrieval costs\n- **Multi-Cloud Cost Tracking** - Unified view across AWS, Azure, GCP\n- **AI API Usage** - Track OpenAI, Anthropic, and other LLM API costs\n- **Executive Dashboards** - Role-based views for CFO, COO, Infrastructure Leaders\n- **Cost Attribution** - Per-model, per-team, per-project cost breakdown\n- **AI-Powered Optimization** - Intelligent recommendations for cost savings\n\n## \u2728 Key Features\n\n### \ud83d\udcca Comprehensive Observability\n- **Real-time Monitoring**: Live metrics for all AI/ML operations\n- **Multi-Cloud Support**: AWS, Azure, GCP telemetry integration\n- **LLM Training Tracking**: GPU utilization, loss curves, cost per epoch\n- **RAG Analytics**: Document processing, embedding generation, retrieval accuracy\n- **API Cost Tracking**: OpenAI, Anthropic, and custom API endpoints\n\n### \ud83d\udcb0 FinOps Intelligence\n- **Cost Observatory**: Centralized cost tracking and analysis\n- **Cost Attribution**: Per-model, per-project, per-team breakdowns\n- **Budget Management**: Set budgets, track spending, get alerts\n- **Optimization Recommendations**: AI-powered cost-saving suggestions\n- **Trend Analysis**: Historical cost patterns and forecasting\n\n### \ud83d\udcc8 Professional Visualizations\n- **VizlyChart**: Built-in visualization library for charts and graphs\n- **Executive Dashboards**: Role-based dashboards (CFO, COO, Infrastructure)\n- **Real-time Charts**: Live updating metrics and KPIs\n- **Custom Reports**: Generate and export custom cost reports\n\n### \ud83d\udd10 Enterprise-Grade Security\n- **Role-Based Access Control (RBAC)**: Fine-grained permissions\n- **IAM Integration**: Identity and access management system\n- **Audit Logging**: Complete audit trail for all operations\n- **Secure API**: Authentication and authorization built-in\n\n### \ud83d\ude80 Easy Deployment\n- **Docker Support**: Containerized deployment ready\n- **Kubernetes Ready**: Helm charts for K8s deployment\n- **Cloud-Native**: Deploy on any cloud provider\n- **On-Premise**: Run in your own datacenter\n\n## \ud83c\udfd7\ufe0f Architecture\n\n```\nopenfinops/\n\u251c\u2500\u2500 observability/ # Core observability platform\n\u2502 \u251c\u2500\u2500 observability_hub.py # Central monitoring hub\n\u2502 \u251c\u2500\u2500 llm_observability.py # LLM training & RAG monitoring\n\u2502 \u251c\u2500\u2500 finops_dashboards.py # FinOps dashboards\n\u2502 \u251c\u2500\u2500 cost_observatory.py # Cost tracking\n\u2502 \u251c\u2500\u2500 cost_reporting.py # Cost reporting\n\u2502 \u251c\u2500\u2500 ai_recommendations.py # AI-powered optimization\n\u2502 \u2514\u2500\u2500 alerting_engine.py # Intelligent alerting\n\u2502\n\u251c\u2500\u2500 vizlychart/ # Visualization library\n\u2502 \u251c\u2500\u2500 charts/ # Chart implementations\n\u2502 \u251c\u2500\u2500 rendering/ # Rendering engine\n\u2502 \u2514\u2500\u2500 core/ # Core utilities\n\u2502\n\u251c\u2500\u2500 dashboard/ # Role-based dashboards\n\u2502 \u251c\u2500\u2500 cfo_dashboard.py # CFO financial view\n\u2502 \u251c\u2500\u2500 coo_dashboard.py # COO operations view\n\u2502 \u251c\u2500\u2500 infrastructure_leader_dashboard.py\n\u2502 \u2514\u2500\u2500 iam_system.py # Access control\n\u2502\n\u2514\u2500\u2500 agents/ # Cloud telemetry agents\n \u251c\u2500\u2500 aws_telemetry_agent.py\n \u251c\u2500\u2500 azure_telemetry_agent.py\n \u251c\u2500\u2500 gcp_telemetry_agent.py\n \u2514\u2500\u2500 generic_telemetry_agent.py\n```\n\n## \ud83d\ude80 Quick Start\n\n### Installation\n\n```bash\n# Install from source\ngit clone https://github.com/openfinops/openfinops.git\ncd openfinops\npip install -e .\n\n# Or with all features\npip install -e \".[all]\"\n\n# Or specific cloud providers\npip install -e \".[aws,azure,gcp]\"\n```\n\n### Basic Usage\n\n```python\nfrom openfinops.observability import ObservabilityHub\nfrom openfinops.observability import LLMObservabilityHub\n\n# Initialize observability\nhub = ObservabilityHub()\nllm_hub = LLMObservabilityHub()\n\n# Register training cluster\nllm_hub.register_training_cluster(\n cluster_name=\"gpu-cluster-1\",\n nodes=[\"node-1\", \"node-2\"]\n)\n\n# Track training metrics\nllm_hub.track_training_metrics(\n model_id=\"gpt-custom\",\n epoch=1,\n step=100,\n loss=0.5,\n gpu_memory_usage=8000\n)\n\n# Get cost summary\ncost_summary = hub.get_cost_summary()\nprint(f\"Total spend: ${cost_summary['total']}\")\n```\n\n### Running the Dashboard\n\n```bash\n# Start the web dashboard\nopenfinops-dashboard\n\n# Or with custom port\nopenfinops-dashboard --port 8080\n\n# Access at http://localhost:8080\n```\n\n## \ud83d\udcda Documentation\n\n- **[Installation Guide](docs/installation.md)** - Detailed installation instructions\n- **[Quick Start](docs/quickstart.md)** - Get started in 5 minutes\n- **[API Reference](docs/api-reference.md)** - Complete API documentation\n- **[Architecture](docs/architecture.md)** - System architecture overview\n- **[Deployment Guide](docs/deployment.md)** - Production deployment guide\n- **[Contributing](CONTRIBUTING.md)** - How to contribute\n\n## \ud83c\udfaf Use Cases\n\n### For Engineering Teams\n- Monitor LLM training job costs in real-time\n- Track GPU utilization and optimize resource allocation\n- Debug cost anomalies and inefficiencies\n- Set up alerts for budget overruns\n\n### For Finance Teams\n- Get complete visibility into AI/ML spending\n- Track cost attribution by team, project, or model\n- Generate reports for stakeholders\n- Forecast future AI infrastructure costs\n\n### For Leadership\n- Executive dashboards with key metrics\n- ROI analysis for AI initiatives\n- Budget vs. actual spending tracking\n- Strategic cost optimization recommendations\n\n## \ud83c\udf10 Multi-Cloud Support\n\nOpenFinOps supports telemetry from:\n\n- **AWS**: CloudWatch metrics, Cost Explorer, EC2, SageMaker\n- **Azure**: Azure Monitor, Cost Management, Azure ML\n- **GCP**: Cloud Monitoring, Cloud Billing, Vertex AI\n- **AI Platforms**: OpenAI, Anthropic, Hugging Face, Custom APIs\n\n## \ud83e\udd1d Contributing\n\nWe welcome contributions! Please see our [Contributing Guide](CONTRIBUTING.md) for details.\n\n### Development Setup\n\n```bash\n# Clone the repository\ngit clone https://github.com/openfinops/openfinops.git\ncd openfinops\n\n# Create virtual environment\npython -m venv venv\nsource venv/bin/activate # On Windows: venv\\Scripts\\activate\n\n# Install development dependencies\npip install -e \".[dev]\"\n\n# Run tests\npytest\n\n# Format code\nblack src/ tests/\nruff check src/ tests/\n```\n\n## \ud83d\udcca Roadmap\n\n- [ ] Kubernetes operator for auto-scaling based on cost\n- [ ] Integration with Prometheus and Grafana\n- [ ] Support for more AI platforms (Cohere, AI21, etc.)\n- [ ] Advanced anomaly detection with ML models\n- [ ] Mobile app for cost monitoring\n- [ ] Slack/Teams integration for alerts\n- [ ] Custom webhook support\n- [ ] Cost allocation tagging system\n\n## \ud83d\udcc4 License\n\nOpenFinOps is licensed under the Apache License 2.0. See [LICENSE](LICENSE) for details.\n\n## \ud83d\ude4f Acknowledgments\n\nOpenFinOps is built with inspiration from the FinOps Foundation principles and best practices from the cloud cost optimization community.\n\n## \ud83d\udcdd Citing OpenFinOps\n\nIf you use OpenFinOps in your research, please cite it using the following references:\n\n### BibTeX\n\n```bibtex\n@software{openfinops2024,\n title = {{OpenFinOps}: Open Source FinOps Platform for AI/ML Cost Observability and Optimization},\n author = {Durai and {OpenFinOps Contributors}},\n year = {2024},\n month = {10},\n version = {0.1.0},\n url = {https://github.com/rdmurugan/OpenFinOps},\n license = {Apache-2.0},\n keywords = {finops, cost-optimization, observability, ai-ml, cloud-cost, llm-monitoring}\n}\n\n\n## \ud83d\udcde Support\n\n- **GitHub Issues**: [Report bugs or request features](https://github.com/openfinops/openfinops/issues)\n- **Discussions**: [Community discussions](https://github.com/openfinops/openfinops/discussions)\n- **Email**: durai@infinidatum.net\n\n## \u2b50 Star Us!\n\nIf you find OpenFinOps useful, please consider giving us a star on GitHub! It helps the project grow and reach more users.\n\n---\n\n**Made with \u2764\ufe0f by the OpenFinOps community**\n",
"bugtrack_url": null,
"license": "Apache-2.0",
"summary": "Open Source FinOps Platform for AI/ML Cost Observability with Intelligent LLM-Powered Recommendations",
"version": "0.2.1",
"project_urls": {
"Bug Tracker": "https://github.com/openfinops/openfinops/issues",
"Documentation": "https://openfinops.readthedocs.io/",
"Homepage": "https://github.com/openfinops/openfinops",
"Repository": "https://github.com/openfinops/openfinops"
},
"split_keywords": [
"finops",
" cost-optimization",
" observability",
" ai-ml",
" cloud-cost",
" aws",
" azure",
" gcp",
" openai",
" anthropic",
" llm-monitoring",
" rag-monitoring",
" cost-attribution",
" visualization",
" dashboard",
" telemetry",
" multi-cloud",
" intelligent-recommendations",
" hardware-optimization",
" gpu-optimization",
" scaling-recommendations",
" auto-scaling",
" llm-powered",
" claude"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "905e5c494c4fa20144eaf1abd7a5717681e735a3c3236a5f3c51b6f9d05cbc8c",
"md5": "2d83263c2d67185362f50218e8e246f5",
"sha256": "5d143cd71dc714be9a2ca9bda8a698001826e61ea0389b89ae007b6c205650db"
},
"downloads": -1,
"filename": "openfinops-0.2.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "2d83263c2d67185362f50218e8e246f5",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 864521,
"upload_time": "2025-10-06T23:59:56",
"upload_time_iso_8601": "2025-10-06T23:59:56.895129Z",
"url": "https://files.pythonhosted.org/packages/90/5e/5c494c4fa20144eaf1abd7a5717681e735a3c3236a5f3c51b6f9d05cbc8c/openfinops-0.2.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "8cd961f5e8f61d74327079a9a5bd4e713a0ece45c13a4982e66149a862317244",
"md5": "52505657a3db96cb61eec3bb9007b92d",
"sha256": "89dff9cc7d2f0364cfaf17daa78e9dc3a8fd98124ba9bf96931a59827ed8c2e4"
},
"downloads": -1,
"filename": "openfinops-0.2.1.tar.gz",
"has_sig": false,
"md5_digest": "52505657a3db96cb61eec3bb9007b92d",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 853711,
"upload_time": "2025-10-06T23:59:58",
"upload_time_iso_8601": "2025-10-06T23:59:58.521611Z",
"url": "https://files.pythonhosted.org/packages/8c/d9/61f5e8f61d74327079a9a5bd4e713a0ece45c13a4982e66149a862317244/openfinops-0.2.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-10-06 23:59:58",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "openfinops",
"github_project": "openfinops",
"github_not_found": true,
"lcname": "openfinops"
}
JSON
