with guest Ben Scharfe, Executive Vice President of AI Initiatives, Altera Digital Health
Season 2: Episode 1
Host of AI Amplified, Dr. Heather Bassett, welcomed Ben Scharfe, Executive Vice President of AI Initiatives at Altera Digital Health, to the show. Ben didn’t arrive at AI leadership through a technical background — he arrived through more than a decade in healthcare finance and operations, including leading Altera’s TouchWorks EHR system before taking on AI initiatives across the company’s ten business units. That path, from CPA and controller through a string of acquisitions and business transitions, turns out to be more relevant to AI leadership than it might first appear: the constant demand to learn a new business, a new client base, a new technology, built exactly the kind of adaptability the AI era now requires of everyone.
Host Dr. Bassett has traced a version of the same arc in her own career — hers ran through medicine and a chance partnership with an early data scientist hire, rather than through finance. The duo agreed this throughline is often the same among AI leaders: the people who end up leading AI initiatives are usually the ones who never stopped treating themselves as students.
What sets Altera apart starts with a person, not a platform, according to Ben. He credits much of his own AI fluency to Dr. Bob Taylor, one of Altera’s AI thinkers, and to a habit he picked up early: setting aside regular time to learn directly from him — a practice he now recommends to any leader working alongside deeply technical people.
That same instinct, building the thinking before building the tool, carries through to Altera’s Care Intelligence platform, whose job is to establish the foundational elements any trustworthy AI system depends on: data quality, reliability, and consistency across disparate sources. The goal isn’t AI as a product — it’s governance as the product, with tailored solutions built for different healthcare settings rather than a one-size-fits-all tool forced onto every organization.
Adoption succeeds or fails on one thing, says Ben: whether AI aligns with the workflows clinicians and staff already use. Altera’s work with large payers and health systems, domestically and internationally, reflects that principle — solving for adoption means bringing payers and providers to the same table, stakeholders whose historical friction has long complicated accountable care efforts. That’s the foundational work organizations most often skip in favor of chasing whatever technology is generating the most hype.
Ben shared that Altera’s approach to governance rests on a simple idea: autonomous agents should inherit their authorization from the access control systems an organization already trusts, rather than operating on a separate and potentially looser set of permissions. From there, autonomy scales with risk — high-acuity clinical scenarios warrant tightly limited AI autonomy, while low-risk administrative tasks can be handed off with far greater independence.
Transparency runs through all of it: AI systems need to explain their reasoning and cite their sources, which means health systems have work of their own to do curating the reference materials and medical guidelines their agents will draw from. Dr. Bassett, speaking from her own experience as a physician, points to why that transparency matters so much with clinical audiences specifically: doctors tend to be an inherently skeptical group, wary of having their workflow disrupted without a clear line of sight into why. Trust must be earned before autonomy can be extended — not the other way around.
Ambient AI’s financial case in clinical settings is real, but it’s rarely direct, Dr. Bassett and Ben agreed. The return shows up through leading indicators rather than a clean line to revenue: better clinician satisfaction and reduced burnout, which in turn drive better retention and patient satisfaction. Notably, Altera’s clinical leadership has resisted framing the technology’s value around seeing more patients — that’s not the positioning providers want, and it isn’t the point. The more meaningful shift, in Ben’s words, is letting clinicians spend more time being healers rather than administrators of care.
Dr. Bassett underscores why that restraint matters: burnout is already driving younger physicians and nurses out of the profession, and framing a new tool as a way to squeeze in more volume only turns a clinician into what she calls “a cog in a wheel” — solving nothing while eroding the very workforce the technology was meant to support.
Nursing organizations have recently raised alarms about AI displacing jobs and about the pace of rollout outpacing readiness — a concern Ben doesn’t dismiss, but reframes at its root. Job losses in healthcare are typically driven by financial strain on health systems, not by AI itself. That reframing leads to a counterintuitive point: AI that improves documentation of medical necessity and more completely captures a patient’s risk profile can ease the very financial pressure that leads to layoffs, by reducing denials and improving reimbursement.
For individuals worried about their own roles, Ben’s advice is consistent with his own path: get comfortable being uncomfortable, partner with people who know more than you do, and look for openings to use AI tools wherever you can — administrative functions are likely to be redefined first.
Taking AI from pilot to production starts with a real, well-defined problem, a clear decision upfront about how success will be measured, and peer-to-peer storytelling — colleagues sharing candid, specific feedback — to build the institutional trust that top-down mandates rarely achieve. On the financial side, Ben offered a word of caution about hyperscaler partners: they’re genuinely helpful, but they’re also the ones selling the “gas.” So, an organization that lets a vendor design its architecture unchecked can end up building something impressive and expensive to run.
His recommendation was to deeply understand the true unit economics of the work before scaling, build in deterministic (non-AI) architecture wherever possible to hold down costs, and put hard limits on tool-calling scope so token spend doesn’t quietly spiral.
Healthcare AI is moving fast, and not all of it lives up to the hype. On AI Amplified, Dr. Heather Bassett, Chief Medical Officer at Xsolis, sits down with industry leaders and AI experts to separate what’s real from what’s noise. Each episode explores the innovations paying off, the challenges still ahead, and the lessons shaping what comes next, all with an eye toward restoring the joy in medicine.
AI Amplified is brought to you by Xsolis, an AI-driven healthcare technology company transforming the payer-provider dynamic and uniting these two crucial stakeholders like never before. Since 2013, Xsolis has been breaking down silos to eliminate friction, waste, and unnecessary manual work in healthcare.
Join us in solving some of the biggest challenges in healthcare. Request a consultation to learn more about Xsolis.