AWS Generative AI Training Course–Building Applications with Amazon Bedrock
Key information
Upcoming sessions
Would you like to schedule this training on a specific date? Contact us by email or via the contact form.
Description of the training: AWS Generative AI Training Course–Building Applications with Amazon Bedrock
This intensive two-day training introduces you to developing applications that harness the power of generative artificial intelligence on Amazon Web Services. You will learn to leverage large language models (LLMs) through AWS managed services, in particular Amazon Bedrock, to design innovative solutions tailored to business needs. The programme covers the full lifecycle of building a generative AI application: from understanding the fundamentals of Machine Learning and foundation models, to implementing complete application architectures including chatbots, automatic summarisation systems and code generators. Participants benefit from numerous hands-on labs carried out directly in the AWS console, ensuring operational, immediately transferable skills.
Also discover our Generative AI and LLMs Training, our Generative AI with Azure Databricks DP-3028 Training, our NVIDIA DLI Building RAG Agents with LLMs Training and our NVIDIA DLI Fundamentals of Deep Learning Training. Also discover our Generative AI and LLMs Training Course, our Generative AI with Azure Databricks DP-3028 Training Course, our NVIDIA DLI–Building RAG Agents with LLMs Training and our NVIDIA DLI Fundamentals of Deep Learning Training.
To feed your generative AI models with data: our AWS Data Engineer Training Course builds the data ingestion and preparation pipelines on AWS.
Good to know
This training includes numerous hands-on labs (60% practice) for better learning. The sessions are guaranteed from a single registrant (except in cases of force majeure). A preliminary interview takes place between the participant and/or a company representative in order to take the participant’s profile fully into account (level, needs, professional context, challenges, etc.).
This training does not prepare for an official AWS certification, but it provides a solid foundation for developers wishing to integrate generative AI into their applications.
Assessment: during the training, the trainer evaluates participants’ learning progress through quizzes, role-play scenarios and practical exercises. Participants receive a certificate of acquired skills at the end of the training.
Objectives of the training: AWS Generative AI Training Course–Building Applications with Amazon Bedrock
By the end of this training, you will be able to:
- Understand the fundamental principles of generative AI and its place in the Machine Learning ecosystem.
- Identify relevant business use cases for generative AI and assess their benefits.
- Plan a generative AI project, incorporating risk management and best practices.
- Use Amazon Bedrock to access foundation models and integrate them into your applications.
- Master prompt engineering techniques, from basic to advanced approaches.
- Develop complete applications combining Amazon Bedrock, LangChain and AWS services.
- Implement text summarisation, chatbot, code generation and RAG architectures.
- Apply security and governance best practices for generative AI applications.
Prerequisites of the training: AWS Generative AI Training Course–Building Applications with Amazon Bedrock
To get the most out of this training, basic technical knowledge of Amazon Web Services (AWS Cloud Practitioner level or equivalent) is recommended. Experience in software development, ideally with Python, is also desirable. General notions of Machine Learning and artificial intelligence concepts are a valuable asset for approaching the hands-on labs with confidence.
Detailed programme of the training: AWS Generative AI Training Course–Building Applications with Amazon Bedrock
Introduction to Generative AI: The Art of the Possible
- Overview of Machine Learning and the place of generative AI in the ecosystem.
- Fundamental principles of generative models and large language models.
- Concrete enterprise use cases: content generation, customer assistance, document analysis.
- Risks, benefits and ethical considerations of generative AI.
- Interactive demonstrations to illustrate the potential of these technologies.
Planning a Generative AI Project
- Key terminology and technical fundamentals of generative AI.
- Positioning a generative AI project within the organisation.
- Structuring steps to frame and launch a project.
- Risk analysis and mitigation strategies.
- Best practices to ensure the success of a generative AI initiative.
Getting Started with Amazon Bedrock
- Introduction to Amazon Bedrock: features, benefits and architecture.
- Available foundation models and selection criteria.
- Typical use cases and pricing models.
- Configuring access and navigating the AWS console.
- Hands-on lab: Configuring Bedrock and using the playgrounds.
Foundations of Prompt Engineering
- Foundation model basics and inference mechanisms.
- Prompt engineering principles: zero-shot, few-shot and advanced approaches.
- Model-specific prompt techniques (Anthropic Claude, Amazon Titan, AI21).
- Detecting and preventing prompt misuse.
- Mitigating bias in generated responses.
- Hands-on labs: Refining text prompts and mitigating image bias.
Amazon Bedrock Application Components
- Overview of the components of a generative AI application.
- Foundation models, embeddings and working with datasets.
- Retrieval Augmented Generation (RAG): principles and implementation.
- Model fine-tuning and complementary application components.
- Securing and governing generative AI applications.
- Hands-on lab: Word embeddings.
Amazon Bedrock Foundation Models
- Catalogue of foundation models available in Amazon Bedrock.
- Using models for inference: key parameters and APIs.
- Invocation methods and integration best practices.
- Data protection, traceability and compliance.
- Hands-on lab: Invoking a Bedrock model for zero-shot text generation.
LangChain and LLM Orchestration
- Optimising LLM performance with LangChain.
- Using prompt templates and building processing chains.
- Structuring documents with indexes and managing memory.
- Sequencing components with LangChain chains.
- Managing external resources via LangChain agents.
- Hands-on lab: Integrating Bedrock with LangChain and contextual prompts.
Architecture Patterns and Use Cases
- Reference architectures for generative AI applications on AWS.
- Text summarisation: extractive and abstractive approaches.
- Question-answering systems and intelligent chatbots.
- Code generation and LangChain agents for Amazon Bedrock.
- Hands-on labs: Summarisation with Claude and Titan, conversational chatbot, code generation, LangChain agent integration.
Course highlights
- Focuses on generative AI with Amazon Bedrock
- Covers foundation models, architectures and use cases
- Builds on hands-on labs on AWS
- Prepares you to deploy GenAI applications in production
FAQ
What is Amazon Bedrock?
Amazon Bedrock is the AWS managed service that gives access, through a unified API, to several foundation LLMs (Anthropic Claude, Meta Llama, Mistral, Amazon Titan, Cohere, Stability AI, etc.) without managing GPU infrastructure. It also provides RAG building blocks (Knowledge Bases), agents, guardrails and fine-tuning. MFE-IT trains you to build production-ready GenAI applications.
Which models are available on Amazon Bedrock?
Bedrock offers a catalogue of leading models: Anthropic Claude (conversational assistant, very long context), Meta Llama, Mistral Large and Mixtral, Cohere Command, Amazon Titan (text and embeddings) and Stable Diffusion for images. The MFE-IT training helps you choose the right model for your use case, cost, latency and privacy constraints.
What is the difference between Amazon Bedrock and SageMaker?
Bedrock is serverless and oriented toward turnkey LLM API consumption (pay-per-token), ideal for generative AI. SageMaker is the full AWS ML platform, for training and deploying your own models. The two combine: SageMaker for full MLOps, Bedrock to enrich your apps with LLMs.
How long is the AWS Bedrock training?
The MFE-IT training lasts 2 days (12 hours), a fully tailored format with a maximum of 3 participants per session. It covers the Bedrock catalogue, prompt engineering, agent orchestration, RAG with Knowledge Bases, Guardrails, monitoring and integration with your applications. 30 days of post-training support included.