Foundation Models and Retrieval Augmented Generation with Amazon Kendra and Bedrock

BEGINNER
130 minutes
5 tasks

This lab demonstrates the integration of foundation models with enhanced search capabilities using Amazon Kendra and Amazon Bedrock. You will explore Retrieval Augmented Generation to improve response accuracy and relevancy through advanced search techniques.

Scenario

A financial advisory company aims to improve their customer service by accurately retrieving information using advanced search combined with generative AI methods. You will help them set up and evaluate this system using Amazon services.

Learning Objectives

  • Understand the concept and business use of Retrieval Augmented Generation
  • Implement Amazon Kendra to enhance search capabilities
  • Integrate Kendra with Amazon Bedrock for model-driven search results

tasks (5)

task 1: Set Up Amazon Kendra and Configure Data Sources

30 min

task 2: Integrate Amazon Kendra with Amazon Bedrock Models

40 min

task 3: Evaluate the Retrieval Augmented Generation Setup

30 min

task 4: Implement Guardrails for Secure Model Operations

30 min

task 5: Present an In-depth Report on System Effectiveness

30 min

Prerequisites

  • Basic understanding of AWS services and cloud environment
  • Awareness of AI/ML basics and retrieval systems

Skills Tested

Applying RAG concepts to enhance AI search capabilitiesConfiguring and securing AI models using AWS servicesDocumenting AI strategy and performance evaluation
    Foundation Models and Retrieval Augmented Generation with Amazon Kendra and Bedrock - Hands-On Lab - CertiPass