How I Built an HR Onboarding AI Agent with Amazon Bedrock, Guardrails, FastAPI, React, and Terraform

Description

In this video, I walk through an AI-powered HR onboarding agent I built on AWS using Amazon Bedrock, FastAPI, React, Terraform, Lambda, DynamoDB, SES, Docker, and GitHub Actions. This project shows how to build a practical internal support agent that can: - answer HR policy questions using retrieval over company documents - send welcome emails using Bedrock action groups and AWS Lambda - create onboarding tasks and store them in DynamoDB - apply Bedrock Guardrails for safer and more controlled interactions - stream responses in real time through a FastAPI backend and React frontend Although this example is focused on HR onboarding, the same architecture can be reused for many other internal support use cases such as IT helpdesk, employee self-service, operations support, and policy assistants. Tech stack used: - Amazon Bedrock Agent - Knowledge Base / RAG - AWS Lambda - Amazon DynamoDB - Amazon SES - FastAPI - React + Vite + TailwindCSS - Terraform - Docker Compose - GitHub Actions GitHub Repo: https://github.com/vishal2505/bedrock-hr-agent Medium Article: https://medium.com/towards-aws/build-an-ai-powered-hr-onboarding-agent-with-amazon-bedrock-s3vectors-fastapi-and-react-bee7f52514f8 If you found this useful, please like, share, and subscribe for more AWS, AI, and real-world engineering builds.