Client Name: Trueestate
Trueestate partnered with DevOps TechLab to enhance real estate data analysis with an AI-powered agentic investigation solution on AWS.
The Challenge
Trueestate needed a more efficient way to investigate property and booking performance and understand the reasons behind changes in business activity.
Traditional analysis required users or technical teams to identify the relevant data, prepare SQL queries, execute multiple queries, and manually analyze the results. Investigating a business question across multiple data points could therefore be time-consuming and dependent on technical expertise.
Key challenges included:
- Manual investigation of property and booking data
- Dependency on technical teams for SQL-based analysis
- Multiple queries required to investigate performance changes
- Difficulty connecting insights across different business data points
- Limited ability to investigate complex questions conversationally
- Delayed access to actionable business insights
Discovery
DevOps TechLab worked with Trueestate to understand its real estate analytics requirements and identify investigation scenarios where an AI agent could autonomously explore data and identify potential causes behind performance changes.
The discovery process focused on:
- Understanding property and booking performance requirements
- Identifying relevant database tables and data relationships
- Understanding the types of questions users needed to investigate
- Identifying multi-step investigation scenarios
- Defining secure, read-only access to production data
- Designing the agent’s planning, analysis, and recovery workflow
Based on these requirements, DevOps TechLab designed the QueryDash Agentic Investigation Platform using Amazon Bedrock and Amazon Nova Pro.
Onboarding
DevOps TechLab onboarded Trueestate onto the QueryDash Agentic Investigation Platform, enabling users to investigate real estate performance through natural-language objectives.
For example:
“Investigate why property bookings declined.”
The AI agent interprets the objective and determines the information required for the investigation. It can discover relevant database structures, generate and execute validated queries, analyze intermediate results, and determine whether additional investigation is required.
The onboarding included:
- Database connectivity configuration
- Amazon Bedrock and Amazon Nova Pro configuration
- Secure database credential management
- Read-only SQL validation
- Conversational investigation context
- Multi-step agentic investigation workflow
- AWS application deployment
- Validation of representative real estate investigation scenarios
Operations & Support
DevOps TechLab provides ongoing operational support for the Trueestate solution and its AWS environment.
The focus is on maintaining application reliability, secure data access, and consistent operation of the AI investigation workflow.
Operations and support activities include:
- Application and AWS infrastructure monitoring
- AI agent execution monitoring
- Database connectivity support
- Query execution and validation support
- Investigation workflow troubleshooting
- Production issue resolution
- Application logging and monitoring
- Secure credential and access management
- AWS infrastructure maintenance
Production database access remains controlled through application-level SQL validation. Write and schema-changing operations are rejected before database execution.
Optimisation & Advisory
DevOps TechLab continuously works to improve the performance, reliability, and effectiveness of the Trueestate agentic investigation solution.
Optimisation and advisory activities include:
- Optimising real estate investigation workflows
- Improving agent planning and multi-step reasoning
- Improving SQL generation and validation
- Enhancing error recovery and replanning
- Reviewing AWS infrastructure performance
- Monitoring application and AI-agent performance
- Reviewing security and access controls
- Advising on additional property and booking investigation scenarios
- Identifying opportunities for further AI-assisted analytics
The reusable architecture allows Trueestate to introduce additional investigation scenarios without building a separate analytics solution for every business question.
Summary
DevOps TechLab helped Trueestate introduce an AI-powered Real Estate Performance Investigation Agent that enables users to investigate property and booking performance using natural language.
Instead of relying solely on predefined reports or manually constructed SQL queries, users can provide a business objective and allow the AI agent to autonomously plan and execute a multi-step investigation.
The agent discovers relevant data, performs analysis, evaluates the evidence, recovers from failed queries, and presents findings that help users understand business performance.
The solution combines Amazon Bedrock, Amazon Nova Pro, AWS compute, managed databases, conversational context, and controlled read-only data access to deliver a scalable and secure agentic analytics capability.
Key Benefits
- Autonomous real estate performance investigation
- Natural-language business questions
- Multi-step data exploration
- Dynamic database discovery
- Evidence-based findings
- Automatic recovery and replanning
- Conversational follow-up analysis
- Controlled read-only production access
- Reduced dependency on manual SQL analysis
Note: Add specific business metrics such as investigation-time reduction or analyst-effort savings only after validating them with Trueestate.
About DevOps TechLab
DevOps TechLab is a cloud and DevOps technology company focused on helping businesses design, deploy, secure, and operate modern applications on cloud platforms.
Our expertise spans cloud infrastructure, DevOps automation, application modernisation, cloud security, managed services, Generative AI, and Agentic AI solutions.
By combining cloud engineering with AI capabilities, DevOps TechLab helps organisations automate manual processes, improve operational efficiency, and build production-ready intelligent applications on AWS.
