Doctor Care Anywhere delivers 60,000+ consultations per month. It has patients, clinicians, insurer contracts, employer integrations, and a decade of operational experience in UK digital healthcare. What it did not have — and chose not to build — is an AI operating system for clinical workflows.
Instead, it selected Tandem Health. This pattern is not unique to DCA. It is becoming the default strategy for healthcare platforms at scale.
The Reality of Healthcare Distribution
DCA's value is in its access layer — the contracts with 1,500+ insurers and corporate clients, the 24/7 clinical workforce, the patient acquisition engine, and the regulatory infrastructure for UK telehealth delivery. Building ambient AI documentation from scratch would require: speech recognition tuned for clinical language, natural language processing for medical terminology across specialties, EHR integration engineering, clinical safety certification (DCB 0129, DCB 160), MHRA compliance, and ongoing model iteration with clinical oversight.
That is a different company. DCA correctly identified that its competitive advantage is care delivery, not AI engineering.
Why Partnerships Dominate
Faster deployment. Tandem is already live in 5,000+ care organisations with 100+ EHR integrations. DCA achieved AI integration by connecting to proven infrastructure rather than spending 18-24 months building from zero. Lower regulatory burden. Tandem is CE-marked under EU MDR, compliant with UK Medical Devices Regulations, ISO 27001 and ISO 13485 certified, and meets NHS DSPT and Clinical Safety standards. Building this compliance stack from scratch takes years. Avoids rebuilding complex ML infrastructure. Tandem's system has processed 375,000+ clinical notes in its Capio Ramsay Santé evaluation alone. That scale of clinical data and model refinement cannot be replicated by a platform starting fresh.
The Pattern Is Accelerating
Tandem × Accurx → NHS-wide deployment reaching 200,000+ clinicians across 98% of GP practices. Tandem × DCA → private virtual care workflows at 60,000+ monthly consultations. Tandem × Humanitas → Italy's leading hospital group. In each case: the healthcare platform brings the distribution (patients, clinicians, contracts), and Tandem brings the AI workflow layer.
The same pattern is visible with Heidi Health embedding into provider groups and Tortus AI deploying across NHS trusts. The healthcare AI stack is separating into distinct layers: an access layer (DCA, NHS platforms, hospital groups), a workflow layer (Tandem, Accurx, Tortus), and an emerging decision layer (clinical decision support, still nascent).
Why In-House AI Often Fails in Healthcare
Data fragmentation across legacy EHR systems. Clinical governance constraints that slow iteration. Regulatory requirements (MHRA, CE marking, clinical safety standards) that demand specialist expertise. The speed of foundation model evolution making in-house models obsolete before they deploy. Healthcare organisations that build AI in-house face all of these simultaneously — while competing with specialist vendors whose entire business is solving these specific problems.
What This Means
Most successful healthcare platforms will be integration hubs, not AI builders. The competitive moat for platforms like DCA is not AI capability — it is distribution, clinical workforce, and payer contracts. The competitive moat for AI companies like Tandem is not distribution — it is technology, regulatory compliance, and clinical validation at scale.
The winners on both sides are those who recognise what they are and partner for what they are not.