Hi HAIFer,
The trend of LLM providers moving closer to the patient layer did not slow down this week, and Perplexity just became part of that too. That front door is becoming the new health AI race. I lay it all out here .
P.S. I went inside the platform of one of the companies shaping the UK ambient AI market. The deep dive goes live tomorrow.
Your shortcut for this week:
Read on for the foresight of the week and the full breakdown of what shaped healthcare AI 👇
Solutions and Launches
Perplexity (US): Launched Perplexity Health, integrating electronic health records and wearable data to provide personalized, citation-backed medical answers. [Link]

Diadia Health (US): Launched an AI causal reasoning platform to help clinicians identify root causes in complex chronic cases, especially endocrine, metabolic, and hormonal conditions. Diadia says it showed 98% concordance with expert clinical judgment and reduced diagnostic trial-and-error by 60%. [Link]
DRAI Health (US): Launched an AI-enhanced remote patient monitoring platform that analyzes continuous patient data to detect early warning signs and support earlier clinical intervention. [Link]
Innovaccer (US): Launched Galaxy UM, an AI-powered utilization management platform that uses natural language processing and optical character recognition to automate prior authorization. [Link]
Google (US): Highlighted Google Earth AI for public health research, using geospatial models and datasets to help identify local clusters of under vaccination and support more proactive public health outreach.
[Link]
C the Signs (UK): Launched an NHS-first AI-enabled case-finding pathway for oesophageal and stomach cancers, using electronic health record data and patient-reported information to identify at-risk patients. [Link]
Heidi Health (Australia): Launched Heidi Remote, a 21g wearable microphone for reliable audio capture in clinical settings, with offline capture, later syncing, and encrypted data. [Link]
Ibex Medical Analytics: Launched Ibex 4, an AI pathology platform with automated scoring for key breast cancer markers. A cited study found it identified 57% more patients potentially eligible for targeted therapy by correcting some HER2 0 classifications. [Link]
Governance, Policy, and Ethics
Kaiser Permanente (US): Mental health workers in northern California staged a one-day strike over changes to patient screening, raising concerns about algorithm-led workflows, delayed care, and possible future AI use in behavioral health. Kaiser denied that AI or clerical staff are making clinical determinations or conducting triage. [Link]
State of Utah (US): Utah’s AI regulatory sandbox was highlighted as a model for healthcare AI oversight, allowing supervised real-world testing while regulators learn how to govern these tools based on evidence rather than assumptions. [Link]
Northeastern University (US): Began developing a universal guide for the ethical use of AI in healthcare, aimed at helping health systems turn ethical principles into practical design, development, and monitoring requirements.
[Link]
World Health Organization (Global): Said the rapid use of generative AI tools for emotional support is outpacing evidence on mental health impact. WHO-backed experts called for stronger assessment, monitoring, and governance, including attention to risks such as emotional dependence. [Link]
Research and AI Advancements
Type 2 Diabetes Speech Screening: Researchers validated a speech-based model for Type 2 diabetes using data from over 21,000 individuals. [Link]
Hybrid Breast Cancer Diagnostics: A peer-reviewed study presented an automated framework combining UNet ultrasound lesion segmentation, radiomics, and machine learning to classify breast tumors. [Link]

Deployment and Evaluation of an EHR-integrated, Large Language Model-Powered Tool to Triage Surgical Patients: Stanford deployed an LLM to triage 6,193 surgical patients within its electronic health records. Achieving 94% sensitivity, the study proved that a human-in-the-loop AI system can safely and accurately augment time-intensive clinical screening. [Link]
Partnerships & Adoption
CommonWealth Purchasing Group (US): Partnered with Attuned Intelligence to give community health centers access to a healthcare-focused voice AI platform. [Link]
Marshall Health Network (US): Partnered with Notable to roll out online scheduling and registration/intake workflows for primary care patients. [Link]
Perplexity (US): Partnered with b.well Connected Health to let users securely connect their health records to AI-powered search and answers. [Link]
Basys.ai (US): Collaborated with ePathUSA on a federal pilot for the Centers for Medicare & Medicaid Services. The project utilizes agentic infrastructure to assist human reviewers in detecting medical fraud and waste. [Link]
Datapharm (UK): Partnered with Siguru AI to integrate approved medicine safety information from emc into promotional review workflows [Link]
HealthMutual Group (Hong Kong): Partnered with Hong Kong Data Ltd. to deploy an AI-driven eTPA platform for cross-border healthcare, including digital claims and cashless care in the Greater Bay Area. [Link]
CARE (Global): Partnered with Surgo Health to pilot an AI-enabled system in the Philippines that captures multilingual feedback from frontline health workers in real time. [Link]
Nortal (Global): Partnered with Amigo AI to help healthcare organizations adopt patient-facing clinical AI agents. [Link]
Bets on the Next Health System
Investments:
Earendil Labs (US): Raised $787 million to advance a sprawling pipeline of biologic medicines for autoimmune conditions and cancer. [Link]
Verily Health (US): Secured $300 million to advance the next phase of its precision health AI strategy. [Link]
Latent Health (US): Raises $80M to close the gap between diagnosis and treatment. [Link]
Hybrid Breast Cancer Diagnostics: A peer-reviewed study combined ultrasound segmentation, radiomics, and machine learning for breast tumor classification. [Link]
Health Universe (US): Raised $6 million in seed funding to expand its HIPAA-compliant enterprise AI platform for deploying and governing autonomous and human-in-the-loop agents across healthcare workflows, including oncology and clinical research. [Link]
Dimer Health (US): Raised $13.5 million in a Series A round to support its virtual patient discharge business. [Link]
Flexzo AI (UK): Raised $12 million (£8.9 million) in a Series A led by Octopus Ventures, with participation from Fuel Ventures, for its agentic workforce platform for healthcare providers. [Link]
Parallel (France): Raised a $20 million Series A led by Index Ventures to expand its AI agents for hospital administration, starting with medical coding. [Link]
Health Lean Analytics (Spain): Secured over €2.1 million to optimize hospital surgical suite operations through data automation and IoT integration. [Link]
M&A:
GE HealthCare (US): Acquired medical imaging software provider Intelerad for $2.3 billion to expand its cloud-first, AI-enabled enterprise imaging offering. [Link]
That’s a wrap for Edition #24 of Health AI Foresight.
My goal with this newsletter is simple - to connect the present, the emerging, and the future of healthcare AI.
See you next week.
- Dr. Aboufandi
📩 Something stand out? Or have a suggestion? Hit reply - I read every message.

