Product · December 10, 2025
Product Launch: UMC Utrecht No-show model now available on the Coforix Platform
Every year, tens of thousands of appointments are lost due to no-shows. UMC Utrecht’s AI model predicts which patients are at increased risk of not showing up. This model is now directly deployable as a service via the Coforix platform. This allows interested hospitals to easily implement the model without worrying about hosting or in-house development.
“With the no-show prediction model, we see that AI has a direct impact on outpatient clinics. By calling patients proactively, we achieve 15–20 percent fewer no-shows and make better use of capacity.”
— Ilse Kant, Head of AI for Health, UMC Utrecht
Why is this relevant?
No-shows are a structural burden for outpatient clinics. Capacity is lost, treatment processes are delayed, and administrative workload increases. Many hospitals send standard SMS and email reminders to all patients, but in practice this is not sufficient to prevent no-shows. Targeted outreach to patients at higher risk of not showing up has proven very effective in Utrecht.
Value-add for healthcare professionals and administrators
- Fewer patients miss their appointments and receive timely care
- Better utilization of scarce capacity and shorter waiting times
- Transparent, ethically responsible use of AI
From open source to scalable service
UMC Utrecht developed the solution in-house and implemented it throughout the hospital. The code was shared as open source. However, in practice, it is difficult for other hospitals to implement and adapt the model for their specific patient populations and workflows. Coforix offers an open platform for AI, where in addition to its own agents, third-party AI solutions are also available. The UMC Utrecht No-show model is the first hospital solution now offered via the Coforix platform.
“With the Coforix platform, we make proven AI solutions directly accessible and scalable for all hospitals. This way, institutions can quickly benefit from tangible impact and, if desired, collaborate on model development in a secure environment.”
— Marcel Alberti, CEO & Co-founder, Coforix
Why via the Coforix platform?
- No need for self-hosting or adjustments; Coforix takes care of everything
- Fully integrated into the existing IT landscape (including data warehouse, EHR) and workflows
- Support for implementation, training, and assistance
- One secure environment with full governance and EU hosting according to GDPR
- Supports multiple medical agents and language models from Coforix and third parties
The solution: a predictive model that works in practice
The AI model:
- Uses only appointment characteristics (no socioeconomic data)
- Reliably predicts which patients are at increased risk of not showing up
- A friendly phone call is usually enough to get most patients to attend their appointment
- At UMC Utrecht, this led to a 15–20% reduction in no-shows
The next step for AI in healthcare practice
The No-show model demonstrates how AI adds direct value to core processes. Together with the recently released (clinically validated) Smart Problem List from Coforix, a scalable platform with medical agents for healthcare emerges. Implement once, use multiple agents.
Want to know more?
- Request a demo or discuss implementation options
- Prefer to host and/or train yourself? Check out the open-source version on the UMC Utrecht GitHub: https://github.com/UMCU-Digital-Health/No_Show
For direct questions, you can contact us at hello@coforix.ai.