At the Portfolio Private Health Forum 2026, biomedical engineer and longevity researcher Bertalan Thuróczy argued that artificial intelligence (AI) will play a central role in achieving longer, healthier lives and in everyday health maintenance. He emphasized that the effectiveness of AI systems depends heavily on consistently built, personal longitudinal datasets.
Personal background and beginnings
Thuróczy is 32 years old. Eleven years ago, while a student in electrical engineering at the Budapest University of Technology and Economics (BME), he was diagnosed with a malignant tumor. He received chemotherapy treatments at the National Institute of Oncology (Kékgolyó utca) and, after successful recovery, decided to redirect his engineering skills from disease management toward optimizing and preserving health.
His first device was a sensor-equipped glove designed to provide a digital health solution for Raynaud’s syndrome, a peripheral circulation disorder for which no effective drug therapy exists. He subsequently moved into biosensors and long-term health data collection.
Longitudinal self-monitoring and specific devices
Thuróczy stressed that continuous and accurate data collection is a prerequisite for improvement. He has monitored his physiology over years using multiple devices:
- a smart ring and other wearables to track long-term sleep trends,
- annual continuous measurements with a Firstbeat EKG-based chest strap that records stress, physical activity and sleep with clinical accuracy,
- a smart scale monitoring weight, body fat percentage, estimated bone density and skeletal muscle mass,
- annual use of a continuous glucose monitor (CGM) to optimize nutrition and to study the impact of dawn cortisol rises on blood glucose and insulin response.
A gut microbiome sequencing revealed dysbiosis caused by his earlier chemotherapy; a targeted antibiotic and probiotic regimen resolved gastrointestinal symptoms that had persisted for years within a few days. He integrated his DNA profile into his system using sequencing results from a U.S. provider (the former CircleDNA).
The Meta Sapiens concept — a “third brain” and local LLM
During his medical doctoral research Thuróczy developed the Meta Sapiens concept: a so-called “third brain” intended to augment biological systems. The term “second brain” is commonly used for the enteric nervous system, so Meta Sapiens denotes the next, digital processing layer.
The system runs in a local, encrypted environment using the Obsidian note-management platform, so data remain on the user’s own hardware. Structured health data are indexed as virtual neurons and synapses, and a large language model (in this case Claude) analyzes and synthesizes them. Thuróczy noted that no single practitioner can comprehensively analyze five years’ worth of detailed physiological data at once, whereas a personalized LLM can synthesize that information and make it actionable.
Practical uses: training plans and psychological patterning
While preparing for marathons, Thuróczy uses AI as a 24-hour assistant alongside a human running coach. The system analyzes five years of Strava data (workouts, swimming, running, stretching), correlates it with blood glucose, gut microbiome status, genetics and aerobic capacity, and proposes personalized training plans.
He digitizes his daily handwritten journal using optical character recognition (OCR) technology based on Gemini and feeds it into the locally running, encrypted model. Over years, the AI can reveal emotional patterns and provide an objective mirror without data leaving the user’s control.
Clinical integration, data context and labor market impact
Thuróczy also warned that purchasing software subscriptions alone is insufficient: without appropriate personal context and multi-year medical history or datasets, monthly licenses are of limited use. He noted that in the United States, Open Evidence already operates large language models that physicians use for clinically validated decision support.
Responding to concerns about the labor market, he said AI will not replace physicians but will integrate into clinical workflows and make care more efficient. According to Thuróczy, the ultimate goal of longevity research is not merely to increase years of life but to improve the quality of the time available and maximize years lived in good health.
Closing remark
Thuróczy based his presentation on personal experience and his own developments, underscoring that durable, clinically useful decision support requires multi-year, detailed personal datasets and local data management.
Cover image credit: Portfolio
(This article was prepared with the assistance of an AI; the final content was edited and verified by our journalist.)
