Halsey Wise, CEO of Lyric – Interview Series

Halsey Wise, CEO of Lyric, brings nearly three decades of executive leadership, board governance, and value-creation experience across healthcare technology, software, and financial services. After serving on Lyric’s board since 2022, he stepped into the CEO role with a focus on scaling the Lyric42 Platform, deepening client and partner relationships, accelerating innovation and AI investment, and driving growth through both organic expansion and strategic M&A. Wise has led six companies as CEO or president, including public-company leadership roles at Intergraph and MedAssets, and has served on the boards of major technology and healthcare organizations including Cotiviti, Cerner, Aspen Technology (AZPN ), Acxiom, Nextech Systems, Lyric, and WellSky. He is also chairman and CEO of Lime Barrel Advisors, a senior adviser to TPG Capital, and co-leads a family foundation supporting scholarships, food security, veterans’ services, and healthcare causes.

Lyric is a healthcare decision intelligence company focused on improving payment accuracy for health plans by applying policy, data, AI, clinical review, and explainable decision support before claims are paid. Its Lyric42 Platform is designed to help health plans coordinate payment decisions across rules, workflows, payment integrity products, and AI-driven recommendations, with the goal of reducing administrative waste, improving first-pass payment accuracy, lowering provider friction, and giving clients greater transparency into why payment decisions are made. The company positions its AI approach as one that augments human expertise rather than replacing it, combining automation, clinical validation, governance, and client control to support more consistent and defensible healthcare payment decisions.

You’ve led and advised major healthcare and enterprise technology companies, including MedAssets, Cerner, Cotiviti, and now Lyric. How has your experience across healthcare IT, analytics, and technology shaped your vision for Lyric?

Across every company I have led or advised, one conviction has only grown stronger: healthcare cannot improve without trust, truth, and transparency at its foundation. These are not soft values. These are the operating conditions on which a better healthcare system depends. Every plan, provider, and member is affected by whether the underlying intelligence can be trusted, whether the information reflects truth, and whether the reasoning behind a decision is transparent enough to withstand scrutiny.

That conviction shapes how I think about Lyric. I joined Lyric’s board in 2022 before stepping into the CEO role, and what I understood then I believe even more strongly now. We at Lyric are building one of the most innovative, impactful, and important healthtech companies.

Lyric is healthcare decision intelligence. Decades of clinical and payment expertise, deep client relationships, and a platform operating on a national scale give us a foundation few companies in healthcare have. Payment is the highest stakes expression of that today, but the broader opportunity is to give people in healthcare the technology and insights to make better decisions, sooner, with confidence, in service of better healthcare for the people and patients it ultimately exists to serve.

I also believe deeply that healthcare decision-making belongs to humans. Clinical judgment, policy interpretation, and accountability cannot be outsourced to a model. What AI can do, and what it should do, is inform and support the people making those decisions. It can surface signals, trends, and anomalies that were previously buried in scale and complexity. It can highlight where something looks inconsistent, where a pattern is emerging, or where a decision needs a closer look. That is AI ‘s role in healthcare: To expand human visibility and strengthen human judgment, not to replace it.

Lyric has spent the last several years rebuilding its technology stack into a cloud-based AI platform supporting payment operations across 190 million lives. What were the biggest architectural or organizational challenges involved in that transition?

Lyric originated as a carve-out from a larger organization. We inherited legacy infrastructure alongside incredible client relationships and decades of operational expertise.

Over the last several years, we rebuilt the technology foundation from the ground up. The challenge wasn’t simply modernizing infrastructure. It was evolving a 30-year payment accuracy business into a modern healthcare decision intelligence platform while preserving explainability, auditability, and trust.

A core principle was to ensure zero business disruption as we modernized our technology. We leveraged the advanced capabilities of Lyric42 to ensure highly accurate migrations, working closely with our health plan clients to create a safe and secure transition.

Architecturally, we moved from static predetermined ordering to real-time intelligence that supports first-pass payment decisions within workflows. At the same time, every outcome had to remain transparent, policy-grounded, and explainable.

As we migrate clients onto the new platform, we’re supporting payment operations across 200 million lives through a unified intelligence architecture that surfaces insights earlier, reduces waste, and supports more consistent decision-making at scale.

You’ve argued that healthcare payments need to shift from post-pay recovery toward pre-pay intelligence. What technological breakthroughs are finally making that transition viable at scale?

The shift from post-pay recovery to pre-pay intelligence is now possible because healthcare finally has the combination of large-scale workflow data, cloud infrastructure, and AI capable of separating signal from noise across highly complex claims environments.

Historically, most systems were built to analyze what happened after payment. Today, we can apply intelligence much earlier within workflows, before errors, disputes, and waste move through the system.

From a capability perspective, Lyric has introduced intelligent selection for prepay review ensuring that claims are not unnecessarily held before payment when timing is critical to both plans and providers. Through our intelligent audit platform, we have also enabled claims that need human review to be processed in hours or days, rather than the weeks or months that traditional processes can require.

These technological breakthroughs, paired with the advanced orchestration within Lyric42, have unlocked healthcare decision intelligence for pre-payment accuracy by combining AI, policy, and expert clinical review to support accurate, explainable payment decisions at first pass. Importantly, this only works at scale when intelligence remains transparent, auditable, and grounded in human accountability.

Much of healthcare still relies on fragmented legacy infrastructure. Why do you believe AI-native systems will fundamentally outperform traditional “systems of record” over the next decade?

Legacy healthcare systems were primarily designed to record transactions and support retrospective audits. They were never built to support intelligence-driven decision-making within workflows.

Healthcare now requires intelligence systems capable of operating at the scale, speed, and complexity modern payment operations demand. The advantage of modern healthcare decision intelligence systems is not simply applied AI. It’s the ability to combine policy, data, AI, and expert review to support more consistent, explainable decisions in real time.

The future advantage will belong to systems that can continuously learn from operational workflows while remaining transparent, explainable, and accountable. Healthcare cannot operate without human expertise, so the goal is not autonomous decision-making. The goal is augmenting human decision-making with intelligence people can trust.

Lyric describes itself as a payment integrity AI company built on decades of clinical expertise. How important is proprietary healthcare data and historical claims intelligence in building defensible AI systems in this sector?

Proprietary claims intelligence and workflow data are foundational in healthcare because generic models alone are not enough to support accurate, explainable payment decisions at scale.

Lyric’s platform supports approximately 200 million lives and is built on decades of operational claims intelligence. That proprietary knowledge and context matters because it reflects how policies are applied, where disputes emerge, how errors occur, and how payment workflows behave in the real world.

The combination of workflow integration, three decades of clinical expertise, and longitudinal claims intelligence creates systems that are not only more accurate, but more explainable and operationally auditable.

In healthcare, intelligence only matters if people can understand, validate, and act on it with confidence.

One of the biggest concerns in healthcare AI is governance and explainability. As AI takes on more payment decisioning responsibilities, what does effective human-in-the-loop oversight actually look like in practice?

For us, human-in-the-loop isn’t a buzzword; it’s the architecture. AI is a tool we use to augment human decision-making.

You can see that in our content authoring work. We built the Studio42 content authoring experience on top of our Lyric42 platform, following the principle that every rule and edit must be deterministic. AI helps our teams review large volumes of government, industry, and research material, identify relevant patterns, and surface recommendations. But the AI does not become the source of truth. The source of truth is grounded in original materials and human clinical expertise.

Every rule and edit is designed to be traceable and auditable. Clinical experts review and validate the work before it reaches production, and every recommendation our AI makes to help our clinical team create content is grounded in original sources and human knowledge. We also keep track of the full history of changes, so it is clear whether a rule or piece of content was built from a specific government, industry, or research source.

That is what effective oversight looks like in practice: AI helps with scale and signal detection, clinical experts validate the work, and the system preserves a clear audit trail from source material to reviewed content. Just as important, our clients retain full control over whether and when to implement the content, rules, or recommendations surfaced through the platform.

In healthcare, the stakes are too high for important decisions to be made solely by a machine or model. At Lyric, we believe human expertise, judgment, and accountability must remain fundamental, not treated as an afterthought, designed into the architecture from the start.

Healthcare waste remains one of the largest cost drivers in the U.S. system. Where do you believe AI can have the fastest measurable impact in reducing unnecessary spending without negatively affecting patient care?

Healthcare accounts for about 20% of U.S. GDP, and 25% of that is waste. A majority of that waste is attributed to system complexity, not fraud or intentional waste.

This is exactly why we built the Lyric42 platform. It shifts the claims process from a pay-and-chase model to prepay/pre-claim intelligence, enabling health plans to get payments right the first time.

A significant portion of healthcare waste comes from administrative rework by having incongruent intelligence across multiple stakeholders. This leads to unnecessary denials and rebills, greater intensity of manual audits, and an increase in provider appeals without meaningfully changing the ultimate payment.

Waste comes from not having the right intelligence at the right time. Lyric provides health plans with decision intelligence and recommendations, validated and backed by clinical knowledge to help identify obvious errors and inconsistencies (coding, units, bundling, medical policy alignment). We do so before payment, which reduces waste and miscoding, cutting down on the back-and-forth of rework, appeals, and adjustments.

Lyric’s platform increasingly integrates automation, predictive analytics, and AI-assisted workflows into claims operations. How close are we to a future where large portions of healthcare payment workflows operate autonomously?

Over the next decade, we’ll see a world where most of the tasks inside payment workflows are automated. But regulatory and compliance requirements will continue to require clear accountability, which means an identifiable human and organizational responsibility. AI, automation, and predictive analytics will streamline tedious daily lifts in healthcare payment workflows, including validating data, spotting patterns, and recommending actions in real time during the first pass of a claim. Sensitive, high-impact, or ambiguous decisions will remain under human review.

Healthcare is fundamentally different from many industries because accountability matters. Payment decisions affect providers, plans, patients/members, and financial outcomes across the healthcare ecosystem.

We are not building a fully autonomous payment AI. We are building healthcare decision intelligence designed to augment human decision-making.

The future is not autonomous healthcare payments. The future is explainable intelligence supporting humans who remain accountable for decisions and outcomes.

Many enterprises are still experimenting with AI in isolated pilots. What separates organizations that are successfully operationalizing AI from those that remain stuck in proof-of-concept mode?

The difference is that successful organizations don’t treat AI as a series of isolated pilots. They treat it as an operating capability and propulsion.

That requires more than a strong model. It requires foundational investments in tooling, monitoring, governance, and workflow design. In our experience, AI only scales when the technology supports the governance processes around it, and when human processes are redesigned around the AI-enabled workflow. You need clear ownership, executive sponsorship, operational alignment, and a constant feedback loop for how AI is performing in production.

That last point is critical. Organizations that operationalize AI do not stop at deployment. They monitor performance, analyze outcomes, identify where the system needs to improve, and continuously improve. AI becomes part of the operating rhythm, not a one-time implementation.

At Lyric, we have seen this pattern in areas like content authoring. We focused on getting one domain right first, with the right tooling, controls, monitoring, and human review. Once that foundation proved successful, we could use the same underlying approach to expand into other domains. That is very different from running a proof of concept that looks good in a demo but never has to survive the complexity of live operations.

The organizations that stay stuck in pilot mode often underestimate that complexity. They may test models on narrow datasets or controlled workflows, but they do not wire AI into real operations with the governance, monitoring, and change management required to sustain it. The organizations that succeed build the foundation first, prove value in a focused area, and then scale with discipline.

Looking ahead, do you believe the long-term winners in healthcare AI will be the companies with the best standalone models, or the ones with the deepest operational integration, workflow data, and network effects across the healthcare ecosystem?

In healthcare, the long-term winners won’t be the companies with the best standalone model in isolation. They will be the companies with the deepest operational integration, the richest workflow data, and the strongest network effects across the healthcare ecosystem.

Models will continue to improve, and many capabilities will become more widely available over time. What is much harder to replicate is the combination of embedded workflow position, longitudinal claims intelligence, and trusted relationships across health plans, providers, and partners.

Lyric’s platform spans approximately 200 million lives and is built from decades proprietary payment accuracy intelligence, with the ability to layer in new capabilities and partners. Because we are embedded early in the payment workflow and operate post-care and prepayment, we have visibility into real transaction flow, not just static datasets. That depth of workflow and outcome data is what helps AI separate signal from noise and move from trend analysis to prediction, recommendations, and action.

Thank you for the great interview, readers who wish to learn more should visit Lyric.