Quick facts
| AI assistant | Mia |
| Users supported | 130,000+ |
| Companies served | 13,000 |
| Countries | 132 |
| Experience | 35 years |
| Availability | 24/7, ten languages |
| Where it runs | dlubal.com + inside Dlubal’s desktop software |
Who is Dlubal Software?
Dlubal Software is a global leader in advanced structural analysis and design software used by structural and civil engineers. The family-founded and run company is headquartered in southern Germany and has sales, marketing, and technical development offices worldwide, including a substantial technical presence in Prague, Czech Republic, and offices in the U.S. Dlubal’s core programs are used in 132 countries. With 35 years of experience, Dlubal serves 13,000 companies and over 130,000 users with its high-precision products.
Prof. Dr. Michael Kraus is an expert in machine and deep learning. He works closely with Dlubal founder, CEO, and technical lead George Dlubal, and together with Dogukan Karatas and the AI team they have been developing Dlubal’s AI strategy and products.
What challenge did Dlubal face?
Dlubal’s support engineers field a wide range of questions—bug reports, licensing issues, billing, and advanced structural-calculation topics. A persistent talent shortage made it hard to cover all time zones, and an internal RAG-style prototype stalled when scaling. The team needed a solution that could
- Provide 24/7, multilingual support without expanding headcount
- Ground every answer in trusted company data
- Integrate deeply with Dlubal’s own software via robust APIs
“We also worked for a week or more on integrating the chatbot into our own software. We had some issues with the sizing and wanted to use the rest method to override, for example, the snippet you provided so that we could use multiple languages. So for us. We saw some of the issues we were always in an open dialogue, especially with Eli. What really is a great plus, and it’s also the opinion of our (CEO) is that you provided constantly really quick and very nice support.”
How did Dlubal solve it with CustomGPT.ai?
- Vendor selection – After evaluating several providers, Dlubal chose CustomGPT.ai for answer quality, security, and API depth.
- Data integration – The team uploaded JSON and PDF manuals, extensive e-learning guides, and an entire sitemap into the no-code platform. Key platform features in use include Data Integration, Anti-Hallucination, Security controls, and Sitemap ingestion.
- Persona tuning – Over an intensive two-week sprint, the AI team fine-tuned Mia’s tone and formatting so that formulas render correctly and answers work in ten languages. A REST override lets the web widget switch languages on the fly.
- Deployment – Mia is embedded on dlubal.com and inside Dlubal’s desktop products, giving users contextual help where they need it most.
Continuous improvement comes from weekly reviews of chat logs, plus the like/dislike signal on each answer. A spike in dislikes sometimes indicates that the user is on an outdated product version—letting the sales team reach out proactively.
What results did Dlubal achieve?
Mia, Dlubal’s incredibly well-trained and knowledgeable AI assistant, is completely operational at dlubal.com, answering technical software and administrative customer questions and providing an impressive civil and structural engineering resource (with citations) for Dlubal’s 130,000+ users and website visitors. Mia has also been embedded into Dlubal software to serve customer needs right where they require the most support.
Dlubal’s next use case for CustomGPT.ai and Mia will be in marketing. However, the close partnership formed between Dlubal and CustomGPT.ai sees the Dlubal team hoping to see even further generative AI functionality from the platform in the coming months, ideally in the form of voice or image developments so, for example, Mia could support structural image rendering in the future, with the right fine-tuning.
“We looked at different vendors and in the end, we chose CustomGPT.ai because for us, it had the best spectrum of quality of answers, ease of use, scalability, and most importantly, API capabilities. We have many internal processes that rely on an automated connection to CustomGPT.ai and its API offers great value.”
| Dimension | Traditional support | Dlubal with Mia (CustomGPT.ai) |
|---|---|---|
| Availability | Limited by time zones and staffing | 24/7, ten languages |
| Where help lives | Email / ticket queue | On dlubal.com and inside Dlubal’s software |
| Grounding | Depends on which engineer answers | Grounded in official company docs, with citations |
| Scale | Constrained by talent shortage | Serves 130,000+ users without added headcount |
| Integration | Manual | Connected to internal processes via API |
Why it worked
- Tight grounding of the LLM in official docs prevents hallucinations.
- Strong API support lets Dlubal weave the bot into its own products and automate internal workflows.
- A continuous feedback loop—weekly tuning plus real-time like/dislike signals—keeps answering quality high.
- Ongoing collaboration with CustomGPT.ai support ensured rapid troubleshooting during rollout.
Conclusion
Mia has become integral to Dlubal’s customer-experience strategy, delivering instant, authoritative answers while freeing skilled engineers to tackle complex issues.
To get in touch with Dlubal, please contact Daniel Dlubal at daniel.dlubal@dlubal.com.
To read other Case Studies like this one, visit https://customgpt.ai/customers/.
Frequently Asked Questions
What is an AI customer support assistant for engineering software?
An AI customer support assistant for engineering software is a specialized AI agent trained on a company’s proprietary product documentation, manuals, and knowledge base. Unlike generic chatbots, it answers domain-specific technical questions, such as structural calculation methods, software configuration, or engineering workflow guidance, with cited, accurate responses grounded in verified source material. Dlubal’s AI assistant Mia, built on CustomGPT.ai, is an example serving over 130,000 structural and civil engineers.
How does CustomGPT.ai prevent AI hallucinations in technical support?
CustomGPT.ai uses a document-grounding architecture that restricts the underlying language model to only generating responses based on ingested source documents. It does not draw on general internet knowledge or speculation. Every answer Mia provides is cited and traceable back to Dlubal’s official documentation, ensuring engineers receive accurate, verifiable information rather than fabricated responses.
Can an AI chatbot support multilingual engineering customers?
Yes. CustomGPT.ai’s platform supports multilingual AI customer support through language detection and REST API-based language switching. Dlubal’s Mia assistant currently serves engineers in ten languages from a single knowledge base deployment, enabling consistent support quality across Dlubal’s 132-country user base without maintaining separate localized systems.
How did Dlubal embed an AI assistant inside its desktop software?
Dlubal used CustomGPT.ai’s REST API to integrate Mia directly into its desktop structural analysis products. This required approximately one week of technical integration work to resolve widget sizing constraints and configure language-override behavior. The result is a contextual in-app AI assistant that delivers guidance to engineers within their active working environment, without requiring them to leave the application.
What types of questions does Dlubal’s AI assistant handle?
Mia handles a broad range of both technical and administrative queries, including structural analysis methodology questions, load case configuration guidance, result interpretation, software bug reporting triage, licensing and activation support, billing and account management, and onboarding assistance for new users. Truly complex or novel issues are routed to human support engineers.
How does Dlubal use AI feedback signals to improve support quality?
Dlubal implemented a continuous improvement workflow combining weekly chat log reviews with a real-time per-response like/dislike rating system. Patterns in negative feedback surface documentation gaps, user confusion points, and signals about outdated software versions in use, enabling both support quality improvements and proactive sales outreach.
How long did it take to deploy the AI assistant?
After vendor selection, the core Mia assistant was configured and deployed within approximately two weeks. This sprint covered data ingestion (manuals, e-learning content, sitemap), persona tuning, formula rendering calibration, and multilingual configuration. The in-app software integration required an additional focused technical effort of approximately one week.
Why did Dlubal choose CustomGPT.ai over other AI vendors?
According to Prof. Dr. Michael Kraus, Dlubal’s AI expert, the decision came down to four factors: the quality and accuracy of AI-generated answers, ease of platform use and no-code deployment capability, scalability to serve a global user base, and most importantly the depth and flexibility of CustomGPT.ai’s API. The API capability was critical because Dlubal required deep integration with its own internal processes and software products.
What business results did Dlubal achieve with AI support automation?
Dlubal achieved 24/7 technical and administrative support coverage for over 130,000 users without expanding its support headcount, a measurable increase in customer satisfaction scores, faster average response times, a significant reduction in repetitive ticket escalations to human engineers, and improved support team efficiency. The feedback loop also generated actionable sales intelligence identifying users on outdated product versions.
What industries can benefit from AI support automation similar to Dlubal’s implementation?
Any technical software company serving a large, globally distributed user base with complex domain-specific support needs is a strong candidate. Industries particularly well-suited include engineering software (civil, structural, mechanical), scientific computing platforms, financial software, legal technology, medical device software, and industrial automation tools, anywhere where answer accuracy is critical and support queries require deep domain knowledge.
