Beachmaster Puts Its Property Records to Work with an AI Property Agent
on AWS

Dudobi built an AI Property Agent on AWS that cut Beachmaster's daily email triage from 2 to 3 hours to under 45 minutes, and lease queries from 15 to 20 minutes to 3 to 5 minutes.

The Challenge

Beachmaster's records were complete, but only as useful as the time someone had to read them. Each incoming email had to be checked against contracts, property records and earlier correspondence before the property manager could decide what to do. On a typical day that meant 20 to 30 emails and 2 to 3 hours of checking, and a single lease query could take 15 to 20 minutes of opening documents until the right clause appeared.

Whether a contractor's quote was reasonable came down to what the manager remembered of similar work, and related requests arriving separately were handled separately, so chances to combine work at the same property were missed. All of this would grow with the portfolio, and the knowledge needed to run it sat with one person rather than in a system the business owns.

The Solution

Dudobi designed and built the Beachmaster Property Agent, running in Beachmaster's own AWS account in the Europe (London) Region. Each day the agent works through the incoming correspondence on its own. For every item it decides which sources and tools it needs, retrieves the relevant lease clauses, maintenance history and earlier emails, links related requests across the day's mail, and recommends a next action. The property manager receives one daily summary by email, which is how they prefer to work, and makes every decision. The agent is read-only: it never replies to anyone or changes a record.

The agent is built with the Strands Agents SDK and runs in Amazon Bedrock AgentCore Runtime, using Anthropic Claude through Amazon Bedrock for reasoning and Amazon Titan for embeddings. Amazon Bedrock Knowledge Bases, with Amazon S3 Vectors, retrieve historical property documents, emails and attachments. AgentCore Gateway is the agent's only route to its tools, over MCP, and Amazon Bedrock Guardrails protect sensitive information.

New email reaches the solution through Microsoft Graph notifications to Amazon API Gateway. AWS Lambda functions validate each notification, queue it in Amazon SQS with a dead-letter queue so nothing is lost, store the email in Amazon S3 and index it in Amazon DynamoDB. Amazon EventBridge triggers the daily run, and Amazon SES delivers the summary. AgentCore Observability, Amazon CloudWatch and AWS X-Ray trace every run and tool call.

Dudobi delivered the work end to end, from discovery with the property manager through design, build, testing and production launch, and continues to monitor and refine the agent based on the property manager's feedback.

The Results

Triage that took 2 to 3 hours a day now takes less than 45 minutes, spent reviewing the daily summary and acting on it. Lease queries that took 15 to 20 minutes are answered in 3 to 5 minutes, and as many as 70% of the agent's recommended actions are taken up without significant change.

Just as important, the knowledge needed to run the portfolio no longer depends on one person's memory. Quotes can be checked against past work, related jobs are grouped, and Beachmaster has a secure AWS foundation on which to build further agents.

About Dudobi

Dudobi is an AWS Advanced Tier Services Partner working with small and medium-sized businesses and nonprofits across the United Kingdom and South Africa. We design, build and run workloads on AWS, from managed infrastructure to agentic AI solutions built on Amazon Bedrock and Amazon Bedrock AgentCore. Dudobi has achieved Managed Service Provider accreditation and competencies including SMB and Nonprofit. Dudobi has been certified to ISO 27001 since 2014.