From the source
- LoanPro worked with AWS Partner AllCloud and the AWS Generative AI Innovation Center to build a multi-agent loan servicing assistant on
Amazon Bedrock and Amazon Bedrock AgentCore, giving servicing agents synthesized borrower context and next-step recommendations.
- Early results show customer call times reduced by up to 15 percent, according to AllCloud's published case study.
- Before the assistant, agents assembled a borrower's picture by hand across payment history, loan terms, prior interactions and available workflow actions, which meant longer calls and service quality that varied with experience.
- The assistant runs on
Anthropic's Claude Sonnet 5.5, Claude Haiku 5.5 and Claude Opus 5.5 on Amazon Bedrock, with model availability varying by AWS Region.
- A composer routes each request to a specialist agent, and agent logic is built with the Strands Agents SDK, an open source framework for building AI agents with tool use.
- Requests enter through LoanPro API v2 on Amazon API Gateway, pass through an AWS Lambda handler and an Amazon SQS queue, and reach tools through a Model Context Protocol server on Amazon Bedrock AgentCore Gateway.
- LoanPro validates each request against the user's permissions and compliance controls and writes the action to its audit trail.






