Generative AI agents to optimize data at Damm

Eliminating duplicate records in hospitality data

Damm faced a challenge in ensuring the quality and consistency of its hospitality establishment data. Duplicate records complicated information management and increased operational costs, as each case required manual review before a decision could be made.

The company needed an AI-based solution that could detect duplicates autonomously, then analyze the available information, and resolve them without manual intervention.

With this initiative, Damm reinforces its commitment to operational excellence, aiming for a unified, high-quality database in the hotels, restaurants, and cafés (HORECA) channel with real-time updates.

Autonomous reasoning with contextual data

NTT DATA designed a proof of concept for Damm using AI agents developed with Microsoft Azure technology. This solution combines automated reasoning with external tools — such as geolocation applications and simulated database queries — to enrich the analytical context before making a decision.

The agent’s workflow includes collecting relevant data, performing contextual analysis and generating a final report. This report, sent automatically by email to a designated recipient, explains both the data used and the decision reached by the system.

Unlike other solutions on the market, this approach stands out for:

• An architecture based on AI agents, enabling autonomous decision-making with advanced reasoning capabilities
• The combined use of Microsoft Azure services and external tools like geolocation to enrich data and provide additional context, enabling more accurate analysis
• Automated delivery of detailed reports setting out the decisions the system make and the reasons for them, offering transparency and a fully autonomous experience for the user

This project was developed in collaboration with Microsoft, NTT DATA's strategic technology partner, using:

• Azure AI Foundry for integrating the language model and executing agent inferences
• Azure Agent Services as the operational core of the system
• Geolocation applications as an external information source to enrich the context of each analysis

AI agent that detects and resolves duplicate entries

The autonomous agent we developed for Damm Group was highly accurate:

• Correct decision-making in 97% of cases
• 92% reasoning accuracy
• 60–70% reduction in time spent improving the database — around 6,000–7,000 hours saved a year

These outcomes confirm the system’s ability to significantly reduce the manual effort previously needed to review each case.

The next step will be to integrate the agent’s workflow, enhanced with GenAI, into Damm’s technology environment. The goal is for the system to operate securely and autonomously, improving the integrity and accuracy of hospitality establishment data within the company’s databases.