iCreatAI News
Skip to content
Trending storyUpdating

LoanPro builds loan servicing assistant with Bedrock AgentCore

1 reports1 sources8 hr agoUpdated

Understand this story

AI overview

Loan servicing company LoanPro, together with AWS partner AllCloud and the AWS Generative AI Innovation Center, built a multi-agent loan servicing assistant based on Amazon Bedrock AgentCore, and the assistant is now running in production. Early results published by LoanPro show customer service call duration reduced by up to 15%. According to LoanPro, the team is developing more AI agents to extend capabilities from loan servicing to more loan process stages such as collections and origination support. AWS notes that the availability of the relevant models varies by AWS Region.

AI overview based on source reports · Updated 7 hr ago

Coverage timeline

Follow the coverage to explore different perspectives.

Nov 18
  1. AWS Machine Learning Blog
    How LoanPro built an agentic AI loan servicing assistant with Amazon Bedrock AgentCore

    LoanPro, with AWS Partner AllCloud and the AWS Generative AI Innovation Center, built a multi-agent loan servicing assistant on Amazon Bedrock and Amazon Bedrock AgentCore using Anthropic's Claude models and the Strands Agents SDK, with a composer agent routing requests to specialized agents. The assistant, embedded in LoanPro's existing web portal, retrieves borrower data through tools exposed via AgentCore Gateway over MCP and generates account summaries in under 15 seconds, with early results showing customer call times reduced by up to 15 percent according to AllCloud's published case study. The team validated the agent against a benchmark of 150 SME-labeled accounts with a 99 percent accuracy target that the agent met, and plans to extend the architecture to collections and origination support.

Activity for this story

Current comparable activity
8
Peak comparable activity
10Nov 18 02:00
Change in the past 24 hours
–

Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.