eks-genai - Design generative AI workloads on Amazon EKS
Provides opinionated guidance for building, training, fine-tuning, and serving generative AI and LLM workloads on Amazon EKS.
Tags
Updated: 2026-09-29Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Select GPU or Neuron hardware
- Design EKS accelerator scheduling
- Choose serving frameworks
- Plan distributed training
- Select ML storage
- Design GenAI observability
- Configure AI gateways
- Optimize GenAI costs
- Apply security baseline
Inputs
- Workload type
- Model family
- Latency requirements
- Cost posture
- Team skill level
- Project timeline
- Serving requirements
- Training requirements
- Storage requirements
Outputs
- Hardware recommendations
- EKS stack recommendations
- Serving architecture guidance
- Training architecture guidance
- Storage recommendations
- Observability recommendations
- AI gateway recommendations
- Cost optimization guidance
- Security guidance
Requirements
- Amazon EKS
- Self-hosted AI or LLM workload
- GPU or Neuron support when applicable