Training: AI on AWS
Hands-on training for teams that want to build, run and govern AI and Agentic AI workloads on AWS:
From choosing the right model to running agents in production.
Training: AI on AWS
Hands-on training for teams that want to build, run and govern AI and Agentic AI workloads on AWS: from choosing the right model to running agents in production.
AI on AWS: from exploration to production
AI on AWS gives you the full stack in one place: frontier models from Anthropic, OpenAI, Amazon and the open-weight ecosystem, managed knowledge bases for your own data, a production platform for agents, and the security, cost and compliance controls you already use for the rest of your workloads. The hard part is knowing which pieces to use, how they fit together, and how to run them without surprises on the bill or the audit report.
Several Clouds teaches the services the way we use them in client projects, with labs on real use cases, so your team leaves able to build and operate AI workloads on AWS, not just talk about them.
What this solves
- From pilot to production
Most teams can build a demo in a day. We show what it takes to ship: architecture patterns, model selection, evaluation, observability and the operational guardrails that keep an AI workload running. - Hands-on
Every topic comes with a guided lab in a live AWS account: calling models through Amazon Bedrock, building a knowledge base over your documents, deploying an agent on AgentCore. - Governance and cost built in
Delivered by AWS practitioners who run AI platforms for clients. Security, IAM, guardrails, cost attribution and compliance are part of every module, not an appendix.
Who it’s for
Executives and product leaders
Which AI use cases are worth funding, what the options are on AWS, where the risks sit, and how to keep cost and compliance under control from day one.
Security, Compliance and FinOps
Guardrails, identity, data residency, cost attribution and token budgets for AI workloads.
Engineering, Data/ML and DevOps
Reference architectures, working code and deployment patterns for models, RAG and agents on AWS.
Packages
Executive Briefing
Duration: 2 hours
Format: lecture with live demos, for leadership and product teams. No prior technical knowledge required
Level: Beginner / Intermediate
Agenda
• The AI landscape on AWS: models from Anthropic, OpenAI, Amazon and the open-weight ecosystem, and when to use which
• What agents change for the business: from chat assistants to agents that act in your systems, and what that means for process, risk and ownership
• Grounding AI in your data: how knowledge bases work and what data readiness they need
• Governance and cost: guardrails, data residency in EU Regions, cost attribution and how to budget an AI workload
• Live demo and use-case prioritization: Amazon Bedrock, a knowledge base and an agent on AgentCore end to end, then the framework we use with clients to pick the first three use cases
Technical Workshop
Duration: 1 day
Format: hands-on labs with supporting lectures, in a live AWS account. For engineering, data/ML, DevOps, security and FinOps teams
Level: Intermediate
Module 1: Models and APIs on Amazon Bedrock
• The Bedrock model catalog and how to compare models for a use case
• Calling models with the OpenAI-compatible and Anthropic-compatible APIs and Bedrock API keys
• Prompt engineering, structured outputs, streaming and cross-Region inference
• Lab: migrate an existing OpenAI or Anthropic integration to Bedrock
Module 2: Knowledge Bases and RAG
• Amazon Bedrock Knowledge Bases: data connectors, parsing, chunking and sync
• Retrieval strategies and agentic retrieval for multi-step questions
• Bedrock Data Automation for documents, images and video
• Lab: build a knowledge base over your documents and query it from an application
Module 3: Agents on Amazon Bedrock AgentCore
• Agent architecture: tools, memory, identity and the Model Context Protocol
• AgentCore Runtime, Gateway, Memory, Identity and Observability
• Building agents with the Strands Agents SDK and deploying them with the AgentCore CLI
• Lab: deploy an agent that uses the knowledge base from Module 2 and an internal API as a tool
Module 4: Governance, security and cost
• IAM, fine-grained permissions and Amazon Bedrock Guardrails
• Cost attribution, token budgets and chargeback for AI workloads
• Evaluation, prompt management and observability in production
• Compliance (GDPR and industry regulation), data residency and responsible AI
• Lab: add guardrails, cost allocation and monitoring to the agent from Module 3
Wrap-up
• Q&A and open forum
• Key takeaways
• Learning resources
What you’ll learn
✓ Choose the right model and AWS service for a use case, and call Anthropic, OpenAI, Amazon and open-weight models through Amazon Bedrock with the SDKs you already use
✓ Build and maintain a knowledge base over company data and ground model answers in it
✓ Design, deploy and monitor an agent on Amazon Bedrock AgentCore with tools, memory and identity
✓ Set up guardrails, permissions, cost attribution and observability so the workload is ready for production and audit
Logistics & Next Steps
Delivery: Virtual or on-site.
Group Size: 4 to 6 participants per session.
Language: English or Bulgarian.
Environment: Labs run in a prepared AWS account; participants need only a laptop and a browser.
Choose one package, both, or a mix. We tailor the agenda and the labs to your use cases and send a proposal.
Request a Workshop
Tell us about your team and the use cases you have in mind. We will come back with a tailored agenda and a proposal.
What our participants say
"I found the hands-on experience with AWS AI tools most valuable. It was interesting to see how easily services like Bedrock can be integrated into real-world applications.”
"Honestly, what I found most valuable about that AWS Al workshop agenda is how hands-on and practical it gets with Amazon Bedrock. Several Clouds' team dive into how to get good results with "prompt engineering" and even cover the nitty-gritty like how to choose the right models and integrate them, which is super useful."
"For a one-day session, I believe they covered everything in sufficient detail and were kind enough to answer all questions we had in detail."
Relevant Success Stories
.png)
.png)
Book a meeting
Ready to unlock more value from your cloud? Whether you're exploring a migration, optimizing costs, or building with AI—we're here to help. Book a free consultation with our team and let's find the right solution for your goals.