Python Carbon Emission Calculator
Budget / SalaryHourly project
TypeFreelance project
LocationRemote
Posted1 hour ago
I need a concise Python project that takes the metrics Azure AI Foundry exposes for each Natural Language Processing agent I deploy and converts them into a clear carbon-emission figure per agent.
Context
• The agents handle Text data only and are limited to a Text classification workload.
• Everything runs on Azure AI Foundry, so the solution must tap into the standard monitoring endpoints or logs that expose runtime details such as vCPU/GPU minutes, memory usage, region and uptime.
Scope of work
1. Collect the relevant utilisation statistics programmatically (Azure SDK or REST).
2. Translate those numbers into kWh consumed and CO₂ equivalent using accepted carbon-intensity factors for the Azure region in use (the Azure Sustainability API or any peer-reviewed dataset is fine).
3. Present a per-agent emission figure and a simple rolled-up report (JSON + CSV).
4. Package everything as a clean, well-documented Python module that I can drop into my existing Foundry workspace.
Deliverables
• Python source code with inline docstrings and a short README explaining setup, required environment variables and how to trigger the reporting job.
• Example output files (one JSON, one CSV) generated from mock or sample metrics so I can verify the format quickly.
• A brief note on the emission factors you chose and their sources.
Acceptance criteria
The script must run with Python 3.10+, output results for each agent in under 30 seconds on a representative log set, and reproduce the same numbers when I feed in identical metrics.
If you have past experience integrating Azure telemetry or sustainability tooling, please highlight it in your proposal.
Context
• The agents handle Text data only and are limited to a Text classification workload.
• Everything runs on Azure AI Foundry, so the solution must tap into the standard monitoring endpoints or logs that expose runtime details such as vCPU/GPU minutes, memory usage, region and uptime.
Scope of work
1. Collect the relevant utilisation statistics programmatically (Azure SDK or REST).
2. Translate those numbers into kWh consumed and CO₂ equivalent using accepted carbon-intensity factors for the Azure region in use (the Azure Sustainability API or any peer-reviewed dataset is fine).
3. Present a per-agent emission figure and a simple rolled-up report (JSON + CSV).
4. Package everything as a clean, well-documented Python module that I can drop into my existing Foundry workspace.
Deliverables
• Python source code with inline docstrings and a short README explaining setup, required environment variables and how to trigger the reporting job.
• Example output files (one JSON, one CSV) generated from mock or sample metrics so I can verify the format quickly.
• A brief note on the emission factors you chose and their sources.
Acceptance criteria
The script must run with Python 3.10+, output results for each agent in under 30 seconds on a representative log set, and reproduce the same numbers when I feed in identical metrics.
If you have past experience integrating Azure telemetry or sustainability tooling, please highlight it in your proposal.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.