Conversation
Highlights

From Health Plan CEO to AI Pioneer: How Siftwell Analytics is Transforming Healthcare Operations

When a health plan merger created an inflection point, Trey Sutten didn’t follow the conventional path of joining the combined organization. Instead, the former Medicaid CFO and health plan CEO saw an opportunity to solve one of healthcare’s most persistent challenges: matching resources with patient needs at exactly the right time.

In a recent episode of Category Visionaries, Trey shared how Siftwell Analytics has raised $5 million to build technology that helps health plans predict the future. But more importantly, he revealed how they’re succeeding in a market where many AI healthcare companies have stumbled.

The Genesis: Starting with Operator Insight

The journey began with a crucial question to fellow C-level executives: what does the market actually need? “When we knew we were going to merge the previous plan, I started to collect the C-level executives and say, ‘Hey, now that we’re figuring out what our next thing is… what is it that the market needs? Where are their holes?'” Trey recalls.

Their collective experience pointed to a fundamental problem: healthcare organizations needed to better align resources with individual needs. But before building anything, they needed to prove their technology could outperform existing solutions.

Validation Before Scale

Rather than rushing to market, Siftwell took a methodical approach:

1. Obtained a sample dataset with millions of lives
2. Assembled data scientists for a competitive analysis
3. Convinced a health plan to provide real, live data for testing

This careful validation paid off. “That was pro bono, and we’ve since converted them to a paying client,” Trey explains. Success with this initial client built confidence in both the technology and their ability to solve unique problems in healthcare.

The AI Adoption Reality Check

Healthcare’s relationship with AI isn’t straightforward. Trey identifies three distinct levels of adoption:

• Level 1: Ad hoc tools for marketing and HR
• Level 2: Back office automation and fraud detection
• Level 3: Patient-to-provider settings and medical decisions

Many companies have struggled with Level 3. “When we’ve seen folks do that in the last couple years, what we found is that there have been some mistakes made,” Trey notes. This has created “trigger shyness” around AI in clinical settings.

Building Trust Through Understanding

At a Digital Health New York panel, Trey witnessed something revealing: founders struggling with basic healthcare terminology. This highlighted a crucial advantage for Siftwell. “Learn the space or learn one space and really focus on it,” Trey advises. “The healthcare industry is plagued by technologists that have really cool mousetraps but don’t understand the complexity of some of the problems that healthcare, particularly managed care organizations, are facing.”

Beyond Predictive Analytics

While Siftwell operates in the predictive analytics space, they’ve created their own category. “This isn’t about how accurate your model is,” Trey emphasizes. “It’s about your model telling me that an individual needs an intervention so that their use of an emergency department or their readmission rates go down.”

This focus on concrete outcomes rather than technical metrics shapes everything from product development to sales conversations. For instance, when helping a client improve cancer screening rates, they didn’t just identify 12,000 members unlikely to participate – they explained why different cohorts within that group might skip screenings, enabling targeted interventions.

The Future Vision

Looking ahead to 2025, Siftwell is building toward automated, personalized care coordination. Trey envisions a system where “I can tell you who’s not going to get a cancer screening, I can contextualize them, including the level of rurality… and if they live in a rural area, I’m shipping automatically a cologuard box versus somebody that lives in a more urban area.”

This vision becomes increasingly important as healthcare organizations face tighter constraints. “People are going to have to get a lot sharper on how they’re spending their dollars,” Trey predicts. “I don’t think that there’s going to be more money going into the next administration.”

Success in healthcare AI isn’t just about better algorithms – it’s about understanding the operational complexities and delivering measurable outcomes. By combining deep industry expertise with advanced technology, Siftwell is showing how to bridge the gap between AI’s potential and healthcare’s practical needs.