First patient-centric language model approach revolutionizes life sciences
- Real-time patient experience data from 100 million sources in 26 languages
- A purpose-built large language model trained on global patient experience data for life sciences
- PatientGPT will replace 6 months of real research with seconds
- A milestone for patient-oriented drug development
Bielefeld, July 11, 2023 | With the announcement of PatientGPT, a new era in drug development begins. Thanks to the interaction of supervised machine learning and large-language models, PatientGPT will deliver critical accuracy – a milestone and a must in life sciences. Drug effects and estimates based on millions of patient experiences can be analyzed and interpreted in seconds, a task that would otherwise take up to 6 months of work. In addition, Europe benefits from strict security measures and anonymization guidelines for sensitive patient data.
For almost eight years, the Bielefeld life science startup Semalytix has been transforming the EUR 1.25 trillion pharmaceutical industry and is continuously delivering advances and key technologies made in Bielefeld. Featuring the world’s first working prototype of an LLM that answers questions cared for
patient experiences, the East Westphalian company is taking another step forward. “We can not only analyze the effect of each drug, but also query it precisely, based on patient experiences and unmet needs worldwide, down to the smallest detail,” says co-founder and Chief Product Officer Janik Jaskolski. “By the end of the year we will have over 50 million patient data points in our patient experience data archive against which our proprietary LLM solution will align. This enables us to gain precise insights into how people with diseases live and generate decisive new insights for patient-oriented drug development,” says cognitive computer scientist Janik Jaskolski.
Keep pushing the boundaries
“PatientGPT truly unlocks the tremendous potential of patient experience data that will help develop new therapies and authentically and continuously answer what patients need most,” says Prof. Dr. Philipp Cimiano, Co-Founder and Chief Technology Officer. The best-known example from recent history is the accidental invention of the powerful drug Viagra in 1998, which was originally a side effect of the tested high blood pressure drug and was put to a new use based on patient experience. Thanks to modern technologies, this type of learning can now be amplified. “With access to over 100 million sources in 26 different languages and the combination of patient experiences from all over the world over the last 10 years, the treasure trove of global patient knowledge already resides in our Pharos platform,” says Janik Jaskolski.
This knowledge is now made accessible thanks to the synergy between the data enriched by supervised machine learning and the incredibly easy interaction of large-language models. It differs from PatientGPT in that its answers are based solely on data already analyzed by Semalytix’s extensive supervised AI framework. This makes the answers more accurate, less prone to hallucinations, and even allows users to check facts on the fly. In addition, the prototype can access real-time data. “Billions of pieces of information about patient-centric experiences reveal completely new possibilities for therapies and medicines,” says Janik Jaskolski. “What are the experiences of patients around the world with drug X, disease Y and what needs to be done to improve their quality of life? PatientGPT aggregates thousands of patient voices and will always give you an answer to this question,” continues Janik.
The new PatientGPT puzzle piece fits perfectly into the AI-based research technology that is already providing insights and evidence for diseases such as depression, breast cancer and diabetes. “We want to dramatically accelerate the speed of therapy development and improve how well new medicines meet patient burdens. We help make the global patient experience as accessible as possible in a compliant and ethical way,” says Chief Product Officer Janik. The numerous case studies, innovation projects and publications already show the upcoming AI-driven medical revolution and underline the important work of the pioneering team from Bielefeld. Semalytix customers include many of the top 20 pharmaceutical companies worldwide.
About Semalytix:
Semalytix is a Bielefeld-based AI technology startup that identifies unmet patient needs and generates a deep understanding of how people experience illness. As the world’s first technology company, Semalytix is committed to improving patient outcomes and developing leading-edge technologies that harness the power of patient experience data to drive informed decision-making in the healthcare industry. Semalytix was founded in 2015 by the managing directors Janik Jaskolski, Philipp Cimiano and Matthias Hartung and currently employs 25 people.
Press contact:
Shima Reza
Global Marketing Director
shima.reza@semalytix.com
