Deploying, Monitoring and Securing LLMs on AWS helps AI/ML professionals get hands-on with generative AI hosting on AWS across two intensive days. Day one opens with an AWS ML services overview, orienting attendees to the managed services available for model work. Attendees then build and deploy models with Amazon SageMaker and scale generative workloads with Amazon Bedrock’s managed foundation model access. They also work with Hugging Face Inference Containers to bring open-source model weights onto AWS infrastructure quickly.
Day two shifts to running those models in production. Attendees start by hosting model weights and other large resources on S3, then focus on optimizing serving throughput so inference stays responsive under load. LLM observability and traceability follow, using OpenTelemetry to trace requests as they move through a serving stack. This skill matters once a pipeline involves multiple chained calls. The course closes with securing models and storage on AWS. Attendees cover access controls and other basic practices that keep model weights and inference endpoints protected.
Four modules per day structure the course, and each spends half its time on hands-on AWS lab exercises. Students work through the AWS services most commonly used for LLM hosting, applying them in realistic lab scenarios. By the end, attendees have the skills and knowledge needed to deploy and secure models effectively on AWS.
Who Should Attend
Data Scientists, Data Engineers, Developers, IT and QA Staff, Technical Managers, AI/LLMOps Engineers
What Attendees Will Learn
Upon completing Deploying, Monitoring and Securing LLMs on AWS, participants will be able to:
- Understand the tools and technologies available on AWS facilitating LLM operations
- Effectively deploy, serve and monitor large language models on AWS
- Take advantage of the basic security features and security best practices for GenAI on AWS
Prerequisites
Participants should have foundational knowledge of AI/ML and some cloud experience.