AI in Assistance
allergoGPT & allergoSMARTDoc – The next generation of AI assistance in allergology
Daniel Schönfelder
Allergology care in Germany faces considerable challenges: high prevalence rates, a shortage of specialists, and short consultation times prevent accurate diagnosis and personalized therapy. AllergoGPT addresses these shortcomings with a novel AI-supported assistance system that uses a federal knowledge graph architecture to efficiently link complex medical data sources and provide evidence-based recommendations. The system will integrate five semantically linked graph databases: a comprehensive allergen ontology, a therapy graph with standardized and individualized immunotherapy preparations, a symptom network with ICD-10-coded manifestations, a literature corpus with publications from the last 10–20 years, and a longitudinal patient database with digital allergy twins. Advanced AI agents and team agents orchestrate multi-graph query processing, link the data deterministically, and deliver auditable, source-based therapy recommendations.