When AI meets the payer: the two-sided fight over healthcare payments
Key points
Key takeaways from a theCUBE + NYSE Wired: MedTech Unplugged interview with Panoramic Health's Sap Sinha and Adonis's Doug Pickett, recorded at the New York Stock Exchange (October 2026):
Healthcare's two-sided problem. AI is being inserted into both sides of an unusual standoff: the clinicians delivering care and the complex machinery that gets that care paid, with Panoramic Health working with nephrologists on kidney care and Adonis applying AI to the revenue cycle.
Back to patient care. Sap Sinha argues physicians have been turned into data operators, spending their evenings answering portal messages and prescription refill requests; AI can summarise charts and triage that workload so doctors can talk to patients again and bureaucracy stops crowding out the humanity of healthcare.
Demand from the frontline. Unlike earlier digital transformations led by tech experts, purchasing pressure is now coming from operating teams and clinicians experimenting with AI themselves and demanding enablement, forcing CIOs and executives to redesign workflows rather than the other way round.
Filling the EHR gap. Legacy systems such as Epic and athenahealth are pushing ahead with expansive, mostly clinical-first AI roadmaps, which leaves gaps in the claims and patient lifecycle that Adonis targets — spotting issues like coding problems or provider credentialing before they become bigger ones.
Fast deployment. Adonis' direct database connection means go-live in about 30 days on athenahealth, the system Panoramic Health runs on, and six to eight weeks on Epic, either layering in its intelligence platform for analytics and alerting or deploying agents to call payers and work payer portals.
Revenue-cycle AI. The platform prioritises where revenue-cycle teams should spend time — aged claims, denials and underpayments — pairing analysis with short-term wins on automation and recommendations rather than relying on analyst hunches.
AI versus AI. With agents active on both the provider and insurer side, Pickett and Sinha argue the rules are undefined: insurers could simply bar agentic negotiation in contracts, and documentation and billing-code requirements were not built for an ambient-AI era that makes clinical notes far more detailed.
Who sets the rules. Sinha cites the Blue Cross and Blue Shield Association reportedly paying about $1bn more because of AI, which he frames not as overpayment but as finally paying appropriately for care, and expects payer-provider friction over contracts and negotiations to ease within three to five years.
Privacy by design. Adonis does not train models on protected health information, de-identifying sensitive data before extracting payer-behaviour insights that inform automations, while HIPAA and business associate agreements keep healthcare buying cycles longer than typical LLM adoption.
Read more: YouTube · theCUBE + NYSE Wired
