The capabilities of the general artificial intelligence platform Phanthy: World model for diabetes risk prediction

Last week, we launched the Phanthy platform,An artificial intelligence platform based on the paradigm AGI route — 'Generative AI + Vertical World Model'. Today, let’s review Phanthy's capabilities in theGlobal Diabetes Risk Prediction Model。

In health management, early risk warnings are particularly crucial. If doctors or health managers attempt to useGeneral AI models, based on users' physical examination reports, to predict the risk of diabetes over the next three years,they often encounter this dilemma:The model may explain indicators like blood sugar and blood lipid levels convincingly, appearing professional, but it often falls short at the critical 'quantitative risk prediction' level—either stating it cannot provide medical diagnoses or giving a vague probability range, lacking verifiable confidence intervals, clear decision-making bases, and error explanations, making it difficult to guide practical interventions.

Faced with the same physical examination report, Phanthy offers a different approach. When health managers upload structured physical examination data, the platform's generative AI quickly recognizes the intent for health risk assessment and precisely dispatches"Global Diabetes Risk Prediction Model"This professional module intervenes in the handling.

This vertical model deeply integrates medical guidelines, large-scale cohort study data, and machine learning algorithms. It does not merely list metrics but outputs a structured risk assessment report within a short time: it provides not only a quantified risk value for developing a condition over the next three years with clear confidence intervals but, more importantly, offers clear decision-making support—for instance, explicitly pointing out that the combination of 'fasting blood glucose at a borderline high level along with central obesity' significantly increases the risk. Additionally, it generates personalized health improvement recommendations, such as prioritizing weight control and regularly monitoring specific indicators.

This once again demonstratesthe 'Generative AI + Vertical World Model'architecture’s versatility in structured decision-making scenarios. Generative AI acts as an intelligent interface for understanding natural language requests and coordinating professional services, while the vertical world model functions as a 'specialized brain,' embedded with domain-specific knowledge and data patterns to perform interpretable and quantifiable in-depth analysis and predictions.

Founded in 2014, Paradigm Intelligence is a publicly listed company and a leader in China's AI + enterprise services sector. After achieving group-level upgrades in 2025, the company's main business covers B-end operations targeting enterprise services under Fourth Paradigm, and C-end operations targeting consumer electronics markets under Phancy.