Open-source sustainability engine for carbon-intensity data, automated CO₂e calculations, DevOps carbon-aware scheduling, and ML-ready datasets.

CarbonOps is a modular sustainability toolkit that automates and standardizes carbon-emission workflows using officially published datasets from DEFRA, IPCC, and global GHG reporting standards.

It provides:

• Dataset ingestion & normalization (CarbonFactor-Parser)
• REST API for CO₂e calculations & factor lookup (CarbonFactor-API)
• CLI + DevOps automation workflows (CarbonOps Toolkit)
• HuggingFace datasets + notebooks + examples
• ML-Ops–friendly data pipelines for future model training

Status: pre-alpha (v0.0.1) — API and internal structure may evolve.

Modern organizations face major challenges:

• Manual or inconsistent carbon reporting
• Scattered datasets and Excel-based workflows
• Limited reproducibility
• Lack of automation
• No ML-driven estimation or validation

CarbonOps standardizes the entire pipeline.

{
  "activity": "diesel",
  "value": 50,
  "unit": "liter"
}

{
  "co2e": 134.5,
  "factor": 2.69,
  "source": "DEFRA 2023",
  "scope": 1
}

✔ Reproducible
✔ Machine-readable
✔ API-ready
✔ ML-Ops compatible

                +----------------------+
                |  DEFRA / IPCC / GHGP |
                |    Raw Datasets      |
                +----------+-----------+
                           |
                           v
                 CarbonFactor-Parser
               (Dataset ingestion layer)
                           |
                           v
                 Normalized Carbon Factors
                           |
                           v
                   CarbonFactor-API
               (REST API & computation)
                           |
                           v
                     CarbonOps Toolkit
       CLI • Workflows • DevOps Carbon-Aware Scheduling
                           |
                           v
               Applications • CI/CD • ML Pipelines

Each module is fully open-source and independently installable, but designed to work together as an integrated sustainability engine.

Dataset ingestion & normalization engine
🔗 https://github.com/ktalpay/CarbonFactor-Parser

REST API exposing standardized emission factors & CO₂e calculations
🔗 https://github.com/ktalpay/CarbonFactor-API

CLI + automation workflows + carbon-aware DevOps utilities
🔗 https://github.com/ktalpay/CarbonOps

git clone https://github.com/ktalpay/CarbonOps.git
cd CarbonOps
pip install -e .
carbonops version

from carbonops import CarbonOpsClient

client = CarbonOpsClient("https://api.carbonops.io")

result = client.calculate(
    activity="diesel",
    value=50,
    unit="liter"
)

print(result)

pip install carbonops
carbonops calculate diesel 50 liter

pip install carbonops-api

docker run carbonops/api:latest

https://huggingface.co/datasets/ktalpay/carbonops-datasets

https://huggingface.co/ktalpay/carbonops-notebooks

https://huggingface.co/ktalpay/carbonops-assistant

from datasets import load_dataset
ds = load_dataset("ktalpay/carbonops-datasets")

• Emission calculation demo
• Dataset exploration
• API integration demo
• ML model prototype (future)

Repo bootstrap + basic CLI

Carbon-intensity adapters + simple scheduler

GitHub Action + per-repo CO₂e reporting

First pilot + documentation site

ML-assisted estimators (missing-value inference)

Carbon-aware CI/CD
Kubernetes carbon scheduling

Issues and PRs welcome.
See CONTRIBUTING.md.

MIT — see LICENSE.

Kürşat Alpay
Senior .NET & AI/ML Ops Engineer
Founder @ FutureOps → https://futureops.co.uk
GitHub: https://github.com/ktalpay
LinkedIn: https://linkedin.com/in/kursat-alpay