Efficient infrastructure
Prioritise cloud providers and services with credible lower-carbon and renewable-energy commitments, while avoiding idle or excessive resources.
Responsibility
Environmental policy
GRI AI is committed to reducing the environmental impact of our operations and the technology choices we make with clients.
Our approach
Better outcomes with fewer wasted resources.
Responsible technology delivery is both an environmental and engineering discipline. We aim to use appropriate infrastructure, reduce avoidable computation and make sustainability part of technical decision-making.
Proportionate AI
Choose models that are appropriate to the task, reduce unnecessary training iterations and favour high-quality, carefully scoped datasets.
Lean data engineering
Evaluate data volume, velocity, variety and quality so pipelines use the right tools and avoid needless processing.
Reuse before retraining
Use transfer learning and existing capabilities where they can deliver the required result responsibly.
Monitor with purpose
Observe model performance and retrain only when evidence shows that it is necessary.
Continuous improvement
Review our practices, involve our team and seek measurable opportunities to reduce environmental impact over time.
Corporate commitment
We seek to comply with applicable environmental requirements, prevent pollution where possible, help our team contribute to better practice and communicate our commitments transparently. We will continue to review our impact and improve this policy as our work evolves.