Our Methodology

Accurate, transparent, and independently verified carbon measurement for cloud and AI infrastructure.

Understanding Cloud Carbon Measurement

Video requires consent to load from the hosting platform.

Our Approach

1. Data Collection

We connect to your cloud provider billing and usage APIs using read-only, least-privilege access. For AI workloads, our SDK captures token counts and model metadata without ever accessing prompt or response content.

2. Carbon Calculation

We combine your actual usage data with real-time grid carbon intensity data from electricity maps and provider-specific Power Usage Effectiveness (PUE) figures. This gives you location-based and market-based emission calculations.

3. Recommendations

Based on your usage patterns, we provide prioritised, actionable recommendations for reducing emissions. These include right-sizing, scheduling, region selection, and architecture changes.

4. Reporting

Generate audit-ready reports aligned with the GHG Protocol (Scope 2 and 3) and the SCI/SCI-AI specifications. Reports include breakdowns by service, region, team, and time period.

Standards Alignment

GHG Protocol

Our calculations align with the Greenhouse Gas Protocol corporate standards for Scope 2 (purchased electricity) and Scope 3 (supply chain) emissions reporting.

SCI Specification

We follow the Green Software Foundation's Software Carbon Intensity specification for standardised, comparable measurement of software emissions.

SCI-AI Specification

Our AI telemetry captures all fields required by SCI-AI: token counts, model identification, compute metrics, and cache efficiency data.

Independent Verification

Our methodology has been reviewed and verified by independent carbon accounting experts to ensure accuracy and completeness.

Want to learn more about our approach?

Read our insights or get in touch to discuss your specific requirements.