• AI scribes reduce burnout: AdventHealth and Mass General report better work-life balance, less paperwork, and more patient time.

​​

• Next step is billing: systems like St. Luke’s already generate $13K extra revenue per physician.

​​

• Mayo leads globally: with 97 AI algorithms in use and 270 in development, plus its platform connecting top hospitals.

• AI scribes began as simple LLM wrappers, but new regulations are raising the bar. Many players won’t survive.

​​

• Smaller firms should target niches, while local champions scale on trust and language; consolidation is coming.

​​

• Scribes are just the entry point: bigger plays in agentic AI and RCM are already underway, with incumbents bundling scribes for free to defend their turf.

• Fintech showed the playbook: unbundle (Stripe, Klarna, Robinhood), then rebundle into digital-first platforms. Healthcare is now on the same path.

​​

• Commure is rebundling via acquisitions (Athelas, Rx.Health, Augmedix, Memora) plus in-house workflow tools.

​​

• Kaiser bets on Innovaccer as the new hospital OS; in Europe, vitagroup and RCM players like Tiplu and Sancare could follow with PE/VC backing.

• Judith Faulkner: the #1 female entrepreneur who built Epic, the world’s leading EHR, from scratch.

​​

• Kaiser Permanente 2003: Epic won against Cerner after showing it could handle KP’s daily flow with zero downtime.

​​

• Epic’s edge: a single unified database and integrated billing/clinical system, delivering unmatched reliability and seamless operations.

​​

• Now in the post-EHR era: open questions remain: integrated vs best-of-breed, the future of hospitals, provider control, Epic’s rigid culture, and whether AI “systems of insights” will overtake EHR “systems of record.”

• Germany’s chance: 250 leading hospitals (incl. Charité, Helios) could accelerate innovation if even half choose modern solutions over legacy HIS.

​​

• Shared data layer: the core should be a vendor-neutral data layer, shifting power away from HIS into a post-EHR era.

​​

• Building blocks: ops platforms (e.g. DNV Imatis) and care pathways (e.g. Careside, UpHill) make the data layer tangible for clinicians.

​​

• Endgame: regional and system-wide data platforms, like Catalonia’s OpenEHR model, Mayo’s federated platform, Kaiser + Innovaccer, or Truveta, show the path, while Bavaria’s failed portal highlights the risks.

• Mayo focuses on clinical & preventive AI: most of its 15 start-ups target predictive health, while Google leans more on ops/admin.

​

• Japan enters the spotlight: 4 of 15 start-ups are Japanese, signaling a rising HealthTech ecosystem.

​​

• Women’s health AI expands: niche solutions now, but platforms like Maven will likely rebundle into full-service care.

​​

• Mayo as blueprint: through Platform_Accelerate and Ventures, it acts as accelerator and VC, reshaping the hospital model.

​​

• Fast pace: since 2022, 45 start-ups joined its platform and 100+ got venture backing.

• Ranking Big Tech in healthcare: by partnership count since 2023: Nvidia leads, Microsoft second, Google third.

 

• Google’s push: from AI in cancer care and genomics (TxGemma) to Pixel Watch 3, Health Connect APIs, Google Wallet health IDs, and 6M AI-powered screenings—Google is scaling fast.

​

• Others in play: Oracle is rebuilding Cerner’s hospital system; Meta hasn’t gone deep yet but Llama 4 could matter as open-source gains traction.

​

• Bottom line: Despite early hype, Big Tech hasn’t disrupted healthcare. Power remains with incumbents, who lean on scale-ups (e.g. Kaiser + Innovaccer) for more control.

• Home care tackles shortages: with a 30% nurse gap ahead, models like DispatchHealth show impact: 1.2M patients treated, 58% fewer ER visits, $1.5B saved.

 

• First wave acquired: Signify Health → CVS ($8B), Care.com → IAC ($500M), ClearCare → Wellsky.

 

• Diverse models: acute care (DispatchHealth), B2B platforms (Medically Home), companionship (Papa), preventive visits (Signify).

 

• Ready for PE: unlike nursing homes, home care enables holistic, end-to-end approaches attractive to investors.

• Only top hospitals can adopt AI agents today: 95%+ lack the infrastructure, leaving adoption to leaders like Mayo Clinic and Kaiser, widening the gap with smaller hospitals.

​

• Smaller hospitals at risk: with poor working conditions and talent leaving, many will shrink into outpatient centers or close, as seen in Germany.

 

• Agents are just the entry point: AI agents will become a commodity; winners will expand deeper into hospital tech stacks. Expect M&A by 2026.

​

• Who wins? Big Tech has scale, but many hospitals prefer smaller players for control. Open-source models may be the real advantage.

• CIOs must act now: AI agents can deliver 10x efficiency, but most hospital IT leaders are stuck maintaining legacy EHRs on tight budgets. Governments should step in with nationwide AI-readiness programs.

 

• Interoperability rethink: some argue agents make standards like FHIR obsolete: AI can structure raw data directly if APIs grant access. Rich, contextual data capture remains critical.

​​

• Bubble or breakthrough? Yes, Agentic AI may be a bubble, but like the internet, not blockchain. Short-term hype may fade, but long-term impact could be massive.

​​

• Open source first: Models like Mistral show open-source AI can rival closed players. In healthcare, community-tested systems may be the safer bet.

• Agreement takes time: AI’s value in radiology is now accepted, but the key is defining when AI outperforms doctors, when humans do better, and where collaboration truly adds value.

​​

• Investor implications: PE-backed radiology chains must rethink strategies—whether to bet on AI-first, human-first, or hybrid models. A hot topic for the HBI Conference in Paris.

​​

• Physician training: Weak AI-assisted results may stem from poor training or “automation neglect,” showing education must adapt.

​​

• Beyond tech and UX: Radiology AI scale-ups must move past algorithms and workflows to redesign operating models and redefine roles.

• D2C clinical is rising: a16z backed Function Health (100+ blood tests by subscription, like Aware in Europe) to predict disease.

​

• Beyond the “worried well”: mostly out-of-pocket today, but Germany may reimburse tests, hinting at preventive care adoption.

​​

• High false positives: at scale, D2C testing risks many false alarms as science lags precision.

​​

• Science too early: gut microbiome and period blood tests remain unreliable, though future potential exists.

• Big VCs are partnering with big Tech and big Pharma.

 

• Kaiser Permanente is building the post-EHR era.
   

• The sale of Dedalus: A new era for hospital IT in Europe?

​

• New developments in AI agents in healthcare emerge every week.

• The silver bullet: Clinicians are in the driver’s seat.

​​

• AI agent ecosystems might leapfrog many existing solutions on the market.

​​

• Hospitals stuck in a rigid, monolithic EHR system risk missing the train.

​

• In 2025, we’ll see whether AI agents truly have transformative potential or if they’re just hype.​

Top 3 events of 2024 (personal opinion):

​

• Hippocratic AI & Tsinghua AI Hospital – AI healthcare agents launched by Hippocratic AI with NVIDIA, while Tsinghua introduced an AI-powered hospital.

 

• AI Scribes by Doctolib & Others – Doctolib and others deployed AI scribes to automate medical documentation.

 

• CompuGroup Goes Private – Leading European EHR provider CompuGroup transitioned to private ownership

Leading German hospital chain Asklepios has just launched Germany's first “virtual hospital wards”.

 

• Remote Patient Monitoring (RPM) required both a rebranding and a comprehensive End-to-End concept. ​

 

• The true benefits of virtual wards are still under discussion. ​

 

• Asklepios' bold move should serve as an inspiration for other German hospitals. ​

 

• This marks the beginning of the unbundling of hospitals, and it’s a positive development.

• Microsoft is the most active of the US Big Tech companies in healthcare. ​

 

• Microsoft Fabric, its unified data store, is particularly promising.

 

• Microsoft is betting on Epic. ​

​

• Microsoft is still pushing for closed-source AI.

• After the unbundling into point solutions, the rebundling into Digital Health Platforms is happening.

 

• More HealthTech M&A is expected in the coming months.

 

• Huma: a big bet on e-consult.

 

• Doctolib is now looking to play a larger role in shaping the HealthTech ecosystem.

Hinge Health, virtual physical therapy for musculoskeletal, and Omada Health, virtual coaching for diabetes and hypertension), have (confidentially) filed the S-1 to prepare for an IPO.

​

• HealthTech IPOs have been a blood bath so far, and DTx have particularly suffered.

 

• However there is hope: 20%+ of US HealthTech going public are still unicorns today.

​

• The inspiration: Evolent Health?

​

• US and Asia have the lion's share of HealthTech IPOs.

• Why is Pfizer launching a D2C platform? Because of macro trends, brand equity, patient access, and bargaining power.

​​

• The success of Hims & Hers must have inspired Pfizer.

 

• Pfizer is willing to build an ecosystem around its indications.

 

• A US approach, not really applicable in Europe.