dormakaba is a global access and security company with thousands of products installed in airports, hospitals, hotels, offices and other complex buildings. Its catalogue spans different generations of hardware, markets and languages.
An AI product assistant tested by 1,000 professionals at BAU
Installers and maintenance teams across dormakaba's global network were losing hours searching through a vast, multilingual product catalogue. Tailor Hub turned that operational problem into an AI assistant grounded in the company's technical documentation and validated it with industry professionals at BAU 2025.
Client
Challenge
Technicians could spend hours locating the right manual, checking compatibility and confirming which parts they needed for an installation. Some products had been created before dormakaba operated in the markets where they were now being used, making the correct documentation even harder to find.
At global scale, every difficult search could translate into lost technical time, a delayed installation, another visit or a support escalation. The company needed a faster way for installers and maintenance teams to access reliable product knowledge in the field.
How the project began
The problem emerged during one of the regular conversations between dormakaba and Tailor Hub. Diego Méndez saw that language models could provide a different interface to the catalogue: technicians could describe their situation in their own words and receive guidance grounded in dormakaba's product information.
dormakaba backed the idea and set BAU 2025 in Munich as its first major test. Installers, architects and technical specialists attending the trade fair would be able to try the assistant with questions drawn from their daily work.
What we built
Tailor Hub designed and built a role-based assistant using GPT-4 through Azure OpenAI, keeping the system within dormakaba's Microsoft cloud environment. The interface runs on Vue and supports German, English and Spanish.
The first mode helped installers retrieve step-by-step guidance from product documentation. The same foundation later expanded to other roles:
Architects can check compatibility, technical constraints and safety requirements.
Contractors can obtain cost estimates and recommendations for specific buildings.
Data analysts can generate reports and visualisations from internal information.
Marketing teams can create content grounded in product specifications.
dormakaba GPT is a perfect example of how an AI project can start with a focused objective and then expand into something much bigger. What began as a tool to save time for installers has grown into a platform that connects roles across the organization, from installers to architects to the marketing team.
Yona Rodríguez - AI & Data Lead at Tailor Hub
Impact
More than 1,000 professionals tested the assistant at BAU 2025. Their questions gave dormakaba an opportunity to validate the system with people who understood the products and the realities of installation work.
The original installer use case has since grown into five role-based modes across three languages. The project also received dormakaba's 2025 Global Tech Innovation Award.