AWS AI & ML
Independent PiSkill directory guide. The original skill remains hosted by AWS Agent Toolkit.
What is AWS AI & ML?
Helps agents select, customize, evaluate, deploy, and operate AI and machine-learning models on AWS, with emphasis on SageMaker workflows and production model lifecycles.
What does AWS AI & ML do?
AWS AI & ML is an AWS Agent Toolkit skill for the full lifecycle of model customization and deployment on Amazon SageMaker. It covers use-case definition, model selection, dataset preparation, fine-tuning, evaluation, deployment, endpoint diagnostics and supporting AWS setup such as IAM, S3 and managed MLflow.
Who is AWS AI & ML best for?
- ML teams customizing models on SageMaker
- Developers selecting a model before fine-tuning or deployment
- Projects evaluating and deploying trained models
- Operations teams diagnosing unhealthy SageMaker endpoints
Common use cases
- Select a base model for a SageMaker project
- Prepare a dataset and run SFT, DPO or other supported training
- Evaluate a trained model before deployment
- Deploy a model and investigate endpoint latency or failures
How does AWS AI & ML work?
The skill routes the request to a focused reference workflow instead of loading every AWS ML topic at once. It can begin with a use-case specification, move into model selection, training or evaluation, and then continue through deployment and production diagnostics while keeping project setup and AWS dependencies explicit.
Key benefits
- Covers planning through production deployment
- Supports multiple model-customization methods
- Includes model evaluation and endpoint diagnostics
- Connects SageMaker model work with IAM, S3 and MLflow setup
Things to know
- It is focused on SageMaker model customization and deployment rather than every AWS ML service
- Training and deployment costs depend heavily on model and hardware choices
- Production quality still requires application-specific evaluation and monitoring
Compatible tools
Frequently asked questions
What does the AWS AI & ML skill cover?
Can the skill help deploy a model without fine-tuning it?
Related skills
AWS Database
Routes AWS database tasks to the right service and workflow, helping agents choose and operate relational, key-value, document, graph, time-series, and caching database technologies.
Hugging Face LLM Trainer
Guides training and fine-tuning language or vision models with TRL or Unsloth while using Hugging Face Jobs infrastructure for execution.
Amazon Bedrock
Guides generative-AI development on Amazon Bedrock, including model invocation, Knowledge Bases, agents, guardrails, AgentCore, model selection, troubleshooting, and related production workflows.
AWS Billing and Cost Management
Analyzes AWS spending, identifies savings opportunities, manages budgets, evaluates commitment discounts, checks pricing, investigates anomalies, and supports cost-optimization workflows.
AWS CDK
Supports authoring, deploying, refactoring, and troubleshooting AWS CDK infrastructure in TypeScript or Python, including stack design, constructs, drift, imports, and safe deployment workflows.
AWS CloudFormation
Guides infrastructure-as-code work with AWS CloudFormation, including template authoring, deployment, troubleshooting, stack behavior, dependencies, and production-safe infrastructure changes.