Optimize over time
You can optimize cost over time by reviewing new services and implementing them in your workload. As AWS releases new services and features, it is a best practice to review your existing architectural decisions to ensure that they remain cost effective. As your requirements change, be aggressive in decommissioning resources, components, and workloads that you no longer require. Consider the following best practices to help you optimize over time. While optimizing your workloads over time and improving your CFM culture in your organization, evaluate the cost
of effort for operations in the cloud, review your time-consuming cloud operations, and
automate them to reduce human efforts and cost by adopting related AWS services, third-party
products, or custom tools (like AWS CLI
Optimize over time
Establish a quarterly generative AI cost optimization cadence that includes:
-
Re-running evaluation benchmarks to validate model price-performance ratios
-
Re-ranking models by business criticality and cost per task
-
Tuning RAG caches and vector retrieval thresholds
-
Pruning inactive embeddings or knowledge bases to reduce silent storage growth
-
Archiving old fine-tuning artifacts to lower storage and inference costs
This helps your generative AI workloads evolve with business demand, maintain cost efficiency, and prevent silent spend creep over time.
Best practice questions
