Clalit and Google Partner for AI Health Assistant
Translated & summarized from Maariv by baba
Clalit Health Services has received a $2 million grant from Google.org to develop an AI health assistant for approximately 5 million patients. The system will use Google Cloud technology and large language models to help patients with appointment scheduling, symptom triage, and managing care, supporting voice interaction in multiple languages and dialects. It will also be accessible via phone for those without internet or smartphones, aiming to improve patient access to healthcare services.
The story in 6 lines · by baba
- Clalit received a $2 million grant from Google.org to develop an AI health assistant.
- The AI assistant will serve approximately 5 million Clalit patients.
- It will support voice interaction in multiple languages and dialects for tasks like appointment scheduling.
- The system will be accessible via phone for patients without internet or smartphones.
- The project aims to reduce medical staff workload and improve patient access to care.
- Clalit plans to release the AI architecture as open-source for global use.
Clalit Health Services announced on Thursday that it has received a two-year grant of $2 million from Google.org, Google's philanthropic arm, to develop an artificial intelligence system aimed at assisting approximately 5 million patients. The system, built on Google Cloud technologies, will enable patients to interact with a digital assistant in their preferred language, receive help preparing for medical appointments, and manage follow-up treatments. This new system is designed to serve as a digital entry point to Clalit's healthcare services, guiding patients through various stages of their medical journey. It will facilitate initial intake and automated symptom triage before doctor visits, with the goal of reducing the burden on medical staff and shortening the time spent on these processes.
A key feature of the development is its voice-based interaction capability, supporting multiple languages and dialects. Clalit explains that text-based digital services are not always accessible to patients who struggle with formal language or applications, hindering their ability to perform tasks like appointment scheduling independently. The system will leverage Google's large language models (LLMs) and Gemini technology, incorporating speech recognition tailored for local dialects and cultural nuances. This will allow patients to complete intake, triage, and appointment scheduling processes in their spoken language.
Beyond smart devices, the development is also intended for patients without internet access or the ability to use apps and SMS. An interactive voice system is planned to operate via landlines and standard mobile phones. Clalit hopes this accessibility will increase the rate of patients independently managing tasks such as scheduling and confirming appointments. The system will be based on a secure cloud infrastructure with data privacy protections, ensuring encrypted interactions and anonymized patient data.
Initially, a pilot program is scheduled, followed by a nationwide rollout. Clalit, which serves over half of Israel's population through 14 hospitals, two mental health centers, and over 1,600 community clinics, plans to eventually release the voice AI architecture and cultural adaptation frameworks as open-source. This would allow other health organizations globally to utilize the model. Prof. Eitan Wertheim is the CEO of Clalit, and Boaz Maoz is the CEO of Google Cloud Israel.
Mentioned
