We forecast healthcare service demand, enabling providers to align their resources precisely when and where they’re needed the most.
What are we building?
For hospitals
Emergency department is the most unpredictable component of a hospital in terms of service demand. High inflow of patients associated with compromised ouflow repeatedly lead to crowding which is an established patient safety issue. Like weather warnings save lives by foreseeing a storm, Tulva AI enables ED stakeholders to predict future crowding while providing up-to-date information about both the past and the present.
For outpatient care
Due to aging populations and difficulty to hire the required professionals, healthcare systems worldwide are struggling to meet the ever increasing demand for service. Demand forecasting in outpatient clinics is crucial for optimizing healthcare resource allocation. By analyzing historical demand statistics and seasonal patterns in combination with related explanatory data, Tulva AI enables clinics to predict patient volumes and efficiently schedule staff.
Team
Antti Roine
Dr. Antti Roine is a physician-entrepreneur who earned his PhD in surgery at 24, pioneering electronic nose technology for disease detection. As CEO of Olfactomics Oy develops advanced healthcare analytics solutions, including an AI-powered smart scalpel that identifies cancerous tissue through gas analysis. At Tulva AI, Antti brings 10 of expertise in commercializing healthcare innovations and translating research findings into solutions.
Eetu Pulkkinen
MSc. Eetu Pulkkinen has majored in Advanced studies in Machine learning and signal processing and he has conducted pioneering research in artificial intelligence early on. Both his bachelor’s and master’s thesis include Tulva AI related research and innovation. At Tulva AI Eetu works as a full stack developer and machine learning engineer and carries the responsibility of translating the research findings to a working prototype.
Niku Oksala
Professor Niku Oksala is a surgeon, researcher, inventor and serial-entrepreneur from Tampere. He works as a chief vascular surgeon at Tampere University Hospital, as a professor of surgery at Tampere University and has co-founded numerous health technology startups. He is a professional visionary and continues to contribute to our mission among his many other responsibilities.
Jalmari Nevanlinna
Dr. Jalmari Nevanlinna combines medical expertise with AI innovation. He is leading the project’s commercialization and research efforts and frontline experience in various healthcare settings drives our mission: to save lives by predictability. He bridges the gap between clinical needs and technical solutions, ensuring the technology delivers real-world impact.
News
Publications
Forecasting mortality associated emergency department crowding
Nevanlinna J, Eidstø A, Ylä-Mattila J, Koivistoinen T, Oksala N, Kanniainen J, Palomäki A, Roine A
J Med Syst 49, 9 (2025)
https://doi.org/10.1007/s10916-024-02137-0Patient Flow Forecasting: Design and Architecture of a Machine Learning-Based Clinical Decision Support System
Pulkkinen E
https://urn.fi/URN:NBN:fi:tuni-202406277428Forecasting Emergency Department Crowding with Advanced Machine Learning Models and Multivariable Input.
Tuominen, J., Pulkkinen, E., Peltonen, J., Kanniainen, J., Oksala, N., Palomäki, A., & Roine, A.
International Journal of Forecasting, xxxx.
https://doi.org/10.1016/j.ijforecast.2023.12.002Early Warning Software for Emergency Department Crowding
Tuominen, J., Koivistoinen, T., Kanniainen, J., Oksala, N., Palomäki, A., & Roine, A
Journal of Medical Systems, 47.
https://doi.org/10.1007/s10916-023-01958-9Forecasting Daily Emergency Department Arrivals Using High-Dimensional Multivariate Data : A Feature Selection Approach
Tuominen J, Lomio F, Oksala N, Palomäki A, Peltonen J, Huttunen H, Roine A
BMC Medical Informatics and Decision Making, 7, 0–37.
https://doi.org/10.1186/s12911-022-01878-7Forecasting emergency department arrivals with neural networks
Pulkkinen E
https://urn.fi/URN:NBN:fi:tuni-202012178976