aws-agentic-ai-customer-support-agent logo

aws-agentic-ai-customer-support-agent

Free

Agentic AI customer support system built with Amazon Bedrock AgentCore, MCP Gateway, RAG, long-term memory, Code Interpreter, Browser, Lambda and API Gateway.

Visit Website
generative-aiagentic-aiai-agentsamazon-bedrockawsaiai assistantautomation

Overview

This project is an agentic customer-support application for an e-commerce scenario, built with Amazon Bedrock AgentCore, Strands Agents, Model Context Protocol (MCP), AWS Lambda, Amazon API Gateway, Amazon Bedrock Knowledge Bases, AgentCore Memory, Code Interpreter, and Browser. The agent routes customer requests to external business tools, retrieves product and policy information, remembers preferences across sessions, calculates loyalty discounts, and accesses web content when needed. It is described as an educational portfolio project, not a production customer-service system.

The agent runs on AgentCore Runtime and loads local and Gateway-provided tools, using customer and session context when handling requests. It can respond conversationally, retrieve information, or invoke an appropriate tool. External actions are demonstrated in the project environment; real-world customer authorization and refund controls would require additional safeguards.

Capabilities include order tracking through an API-backed Gateway tool, refund operations via Lambda-backed Gateway tools, retrieval-augmented generation using a search_knowledge_base tool, long-term memory with AgentCore Memory, loyalty calculations through AgentCore Code Interpreter, and web content access via AgentCore Browser. The project includes automated tests and deployment evidence.

Pricing

Free

Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.

Visit Website

Prices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.

Use Cases

  • Teams use the agent to track order status and shipping details.
  • Teams use the agent to initiate refunds, look up refund status, and generate return labels.
  • Teams use the agent to answer product and policy questions via knowledge-base retrieval.
  • Teams use the agent to remember customer preferences across sessions.
  • Teams use the agent to calculate loyalty discounts and points balances.
  • Teams use the agent to retrieve information from external web pages.

Features

Build agentic customer support flows

Use Amazon Bedrock AgentCore

Retrieve information with RAG

Maintain long term memory

Access tools through MCP Gateway

Use Code Interpreter and Browser

Connect through Lambda and API Gateway

Reviews

No reviews yet. Be the first to share your experience!