Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
PredictWatch is a Polish mental health technology company positioned as an “always-available mental health” tool. Its core focus is addiction management: it collects users’ daily behavioral data through smartphones, smartwatches, and other devices, uses machine learning to predict the intensity of addiction cravings and relapse risk, and attempts to provide interventions to reduce that risk. Products/projects listed on the official website include RiskSense, Cannabe, Nałogometr, as well as earlier offerings such as Guardian, Researcher, and Connector.
Based on the main site content, PredictWatch’s AI capabilities center on “machine-learning-based relapse risk prediction.” The company has received multiple EU R&D grants to validate methods for predicting and reducing addiction relapse risk in real-world conditions, covering opioids, stimulants, alcohol, and mixed addictions, while also expanding into cannabis, gaming, pornography, and other behavioral or substance-related issues. Another R&D direction is a serious therapeutic game based on artificial intelligence and MBRP (Mindfulness-Based Relapse Prevention), planned for release as an iOS/Android mobile app.
The early Guardian product targeted people with addictions who wanted to maintain abstinence, Connector was designed for psychotherapists, and Researcher was aimed at researchers. The company later moved functionality from smartwatches to more widely used smartphones and developed product lines such as Nałogometr and Cannabe. However, the main website content does not disclose the current commercialization status, subscription pricing, enterprise purchasing model, payment methods, or trial policy.
The strengths are its highly vertical focus, addressing relapse prediction as a key pain point in addiction treatment; its product design considers patients, therapists, and researchers; and its substantial EU project funding demonstrates continuity in R&D. The drawbacks are also clear: the website does not provide model accuracy, clinical outcomes, medical regulatory credentials, privacy compliance details, or API information; some descriptions date back to the 2018–2020 period, so the boundaries of currently available products need further confirmation.
It is better suited for addiction rehabilitation institutions, psychotherapists, addiction research teams, and individual users willing to use tools in English or Polish. For Chinese users, Chinese-language support is not disclosed, and access from mainland China is unknown. Because sensitive health data is involved, users should carefully verify the privacy policy, data storage location, and medical compliance before using it.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on predictwatch.com official site.
predictwatch.com is an Poland Health provider. TG4G tracks its product information, an overall rating of 6.0/10, and a China-accessibility score of Limited (proxy recommended). Click "Visit Official Site" to reach predictwatch.com directly.