Specialist Training
AWS AI Engineering Course
Learn to deploy ML models, build LLM applications and operate tool-using agents on AWS. Work with Amazon SageMaker AI, Amazon Bedrock and Amazon Bedrock AgentCore, then connect the workloads to identity, evaluation, observability and release controls.
- ·Choose it when you already have programming foundations and need practical AWS implementation.
- ·Use the course choices below to find broader MLOps, LLMOps or production AI operations training.
At a glance
- Focus
- AWS implementation for ML, LLMs and agents
- Working stack
- SageMaker AI, Bedrock and AgentCore
- Project
- An AWS AI service with evaluated releases
- Starting point
- Python, APIs, Git and basic containers
Who Should Take This AWS AI Course?
This course is for ML engineers, backend developers and platform engineers who want to turn AWS AI services into working systems. It suits teams choosing between managed model endpoints, hosted foundation models and an agent runtime.
You should be able to write Python, call an API, use Git and understand a basic model-evaluation workflow. AWS setup, permissions and resource checks are part of the implementation work.
If you need programming and AI application foundations first, explore the AI Engineering Course.
AWS AI Engineering Syllabus
01Set up the AWS workload
Accounts, IAM roles, regions, service access, S3 artifacts and container images in ECR. Establish budgets, logging and resource-cleanup steps before running the labs.
02Build the SageMaker AI lifecycle
Run a small training workflow, track its configuration and metrics, register a model and deploy an endpoint. Connect pipeline stages to an explicit evaluation and promotion decision.
03Build an evaluated Bedrock application
Choose an accessible foundation model, call it from an application and add retrieval over a controlled document set. Examine Knowledge Bases, guardrails, quality evaluation, quotas and latency.
04Operate an agent with AgentCore
Deploy a bounded tool-using workflow using AgentCore Runtime. Examine identity, session state, tool access through Gateway and operational tracing. Keep model inference, agent orchestration and runtime responsibilities distinct.
05Connect release controls and observability
Version application configuration, run evaluation checks before release and inspect CloudWatch logs and metrics. Trace model and tool calls, handle throttling and design an explicit rollback path.
06Choose managed or Kubernetes serving
Compare SageMaker AI endpoints, Bedrock-managed inference and containers on EKS. Use an architecture walkthrough for KServe on EKS, including operational ownership, scaling, network boundaries and cost drivers.
Build an AWS AI Service with Release Evidence
Use one bounded support workflow, developed through three checkpoints:
ML endpoint.
Train and register a small ticket-classification model, then expose a prediction endpoint with a repeatable evaluation report.
Knowledge application.
Add a Bedrock-backed assistant over a small document set. Record retrieval evidence, answer quality, latency and model usage.
Controlled agent action.
Run a tool-using workflow on AgentCore with explicit permissions and traceable execution. Keep the action bounded, such as looking up a ticket in a test system.
Project artifacts: deployment code, workload architecture, evaluation results and a runbook covering permissions, release, rollback and resource cleanup.
Review focus: selecting the right service, defining workload boundaries and proving that a release can be evaluated and operated. The project is a focused implementation; the EKS comparison is an architecture exercise rather than a second full platform build.
Choose AWS Specialization or a Broader Operations Course
Implement ML, LLM and agent workloads using AWS services
This AWS AI course
Learn the complete ML lifecycle across data, training, deployment and monitoring
MLOps CourseGo deeper into LLM evaluation, serving, reliability, RAG and agent operations
LLMOps CourseStudy ML, LLM and agent operations together in SCAI's broader program
AIOps CourseFor the corresponding Azure implementation, see Microsoft Azure AI Engineering. Self-hosted inference and the broader LLM lifecycle remain part of the existing LLMOps Course linked above.
Training details
AWS AI Course Syllabus, Fees and Training Options
Request the syllabus and current training options for AWS AI Engineering. Tell us whether your priority is SageMaker AI, Bedrock applications or agent runtime operations. The team will confirm course fit, delivery format, schedule, fees and AWS access requirements before enrolment.
By submitting, you agree to be contacted about this program. See our privacy policy.