Developing AI agents for business support presents unique challenges that many organizations face when trying to automate routine HR tasks. Works Human Intelligence (WHI) develops, sells, and supports the integrated HR system “COMPANY” for major Japanese corporations and public interest corporations.

In this post, we share how the AWS Generative AI Innovation Center (GenAIIC) collaborated with Works Human Intelligence (WHI) to build two AI agents using Amazon Bedrock AgentCore. We discuss the challenges encountered and the solutions that reduced costs by up to 97% while improving operational efficiency.

Customers using HR systems must respond to numerous situations, such as organizational changes, revisions to HR systems, and updates to employee information. For organizations facing similar challenges with HR system operations, AI agents can significantly reduce workload and improve productivity. When WHI embarked on building products using AI agents, several challenges arose. To resolve these issues, we at GenAIIC worked closely with the WHI team to provide new perspectives and support in creating a high-quality product.The scope of this project covers two AI agents designed to support the work of operational departments. The Commuting Allowance Agent handles the approval of commuting allowance applications that arise during events like moving. The Browser Operation Agent “COMPANY” on behalf of the customer. We discuss the challenges and solutions for these two agents in the following sections.

This agent automates the approval of commuting allowance applications, which is a routine task that arises during events like employee relocations.

The Commuting Allowance Agent supports the routine task of approving commuting allowance applications. WHI was already proceeding with a proof of concept (PoC) using LangGraph, Amazon Elastic Container Service (Amazon ECS), and AWS Fargate. However, because Amazon Bedrock AgentCore was released during development, the team began considering a migration. WHI wanted to work with us to build a solution with AgentCore that would realize an AI agent integrated with “COMPANY”. Additionally, they wanted to migrate to an integrated multi-agent environment and implement authentication and authorization using AWS Fargate and Amazon Cognito, which were currently under development.

The Commuting Allowance Agent was being developed using LangGraph and Amazon ECS, but the team had concerns about the monolithic configuration where everything ran as the same Amazon ECS task. Therefore, we worked together to change the architecture so that sub-agents are launched individually on the AgentCore Runtime. Because multi-tenancy support was required, we decided to manage tenants using Amazon DynamoDB and Amazon Cognito to maintain the flexibility for WHI to build and manage it.