Siftwell Analytics’ Vision for Agentic AI: Automatically Shipping Cancer Screening Kits Based on Rurality Data

 

Siftwell Analytics’ Vision for Agentic AI: Automatically Shipping Cancer Screening Kits Based on Rurality Data

Most healthcare AI stops at prediction. It tells you which patients are high-risk, unlikely to comply, or headed for expensive interventions. Then it hands you a list and wishes you luck.

In a recent episode of Category Visionaries, Trey Sutten, CEO and Co-Founder of Siftwell Analytics, a healthcare technology company that’s raised over 5 million in funding, described a radically different future—one where AI doesn’t just predict problems but autonomously solves them. “I imagine a future, Brett, and we’re building it where I can tell you who’s not going to get a cancer screening. I can contextualize them, including the level of rurality. I can spin up care plan associated with that member, that individual member. And if they live in a rural area, I’m shipping automatically a cologuard box versus somebody that lives in a more urban area. And I’m automatically setting up the appointment and doing the calls both to the provider and the patient to broker that exchange.”

This isn’t theoretical futurism. It’s the product roadmap Siftwell is actively building today.

Why Predictions Alone Aren’t Enough

Trey’s vision for agentic AI emerges directly from his frustration as a health plan operator. Predictive analytics has been available in healthcare for years. Health plans have lists upon lists of high-risk patients who need interventions. The problem isn’t knowing who needs help—it’s having the operational capacity to act on that knowledge at scale.

“The motivation really is to go from big lists that are rank ordered to bite sized chunks that people can understand, marshal resources around and really get results,” Trey explained. But even bite-sized chunks require human decision-making, resource allocation, and coordination across multiple systems and stakeholders.

This creates a fundamental bottleneck. A health plan might identify 12,000 members unlikely to get cancer screenings. Even with perfect predictions and contextual insights, someone still needs to decide whether to ship screening kits, schedule appointments, arrange transportation, or deploy care coordinators. Each decision requires human judgment, time, and coordination.

Agentic AI eliminates this bottleneck by moving from insight to action autonomously.

The Agentic AI Architecture

What Trey describes isn’t a single feature—it’s a fundamental reimagining of how healthcare AI systems work. The architecture has three distinct layers, each building on the previous one.

Layer 1: Contextual Prediction

The foundation remains predictive analytics, but with crucial context. Siftwell doesn’t just identify patients unlikely to get screened—it understands why they won’t get screened and what factors influence their behavior.

When analyzing cancer screening compliance, the system might segment populations by rurality, transportation access, distance to facilities, affordability concerns, and historical patterns. This contextual understanding becomes the input for automated decision-making.

Layer 2: Dynamic Care Planning

The second layer automatically generates individualized care plans based on predicted behavior and contextual factors. “I can spin up care plan associated with that member, that individual member,” Trey said, describing functionality that’s already in development.

These aren’t generic care plans pulled from a template library. They’re dynamically created based on each patient’s specific situation, barriers to care, and likelihood of responding to different intervention types.

Layer 3: Autonomous Execution

The third layer is where agentic AI becomes genuinely transformative. The system doesn’t just recommend actions—it executes them.

For a rural patient unlikely to get screened, it automatically ships a Cologuard kit to their home. For an urban patient with transportation access, it books an appointment at a nearby facility and initiates calls to both the provider and patient to coordinate the visit.

“That’s the future that I want to build and that we’re already building,” Trey emphasized.

Why Agentic AI Works in Healthcare Now

Healthcare has historically been conservative about automation, especially in patient care. Trey acknowledged this directly, noting that in clinical utilization management, “some folks have gotten in trouble and I think there’s a bit of a snapback or trigger shyness that is manifest from some of those experiences.”

But agentic workflows for preventive care occupy different territory. Automatically shipping screening kits or scheduling appointments isn’t making clinical decisions about treatment—it’s removing operational friction from evidence-based preventive care protocols.

The technology has also reached an inflection point. “Every week is like 10 years in the world of artificial intelligence,” Trey observed. “We’re super fortunate to have a Co-Founder in Eben that can stay on top of this stuff and is actually fascinated by it, which so he’s constantly in the lab.”

This rapid evolution of AI capabilities, combined with healthcare’s growing comfort with automation in non-clinical workflows, creates the window for agentic AI adoption.

The Operational Impact for Health Plans

For health plan operators, agentic AI solves a resource allocation problem that’s only getting worse. As Trey noted when discussing the healthcare landscape heading into 2025: “I think people are going to have to get a lot sharper on how they’re spending their dollars. I don’t think that there’s going to be more money going into the next administration. And so health plans need to figure out how they do more with what they’ve got.”

Agentic workflows dramatically expand what’s possible with existing resources. Instead of care coordinators manually triaging 12,000 at-risk members, the system automatically executes appropriate interventions for thousands while surfacing only the complex cases requiring human judgment.

This isn’t about replacing care coordinators—it’s about letting them focus on cases where human expertise actually matters rather than burning time on logistical coordination that software can handle.

Building Agentic AI: What’s Different

Creating effective agentic workflows requires capabilities most healthcare AI companies don’t have. It’s not just about better models—it’s about integrating across multiple systems and executing real-world actions.

Siftwell’s foundation as former operators gives them crucial advantages here. “We’re taking really deep and battle tested managed care experience, combining it with these advanced Technologies,” Trey explained. Understanding operational workflows from the inside makes it possible to automate them effectively.

The technical architecture also matters. Siftwell already uses multiple types of AI beyond basic prediction. “We use machine learning to make predictions and then we really dig into those predictions using a couple other types of artificial intelligence to explain why those individuals are being identified,” Trey noted.

Adding agentic capabilities means orchestrating these AI systems into workflows that can make decisions and take actions autonomously—a significantly more complex engineering challenge than building prediction models.

The Agentic Workflow Expansion

Cancer screening is just the starting point. The same architectural approach applies across healthcare operations.

Medication adherence workflows could automatically detect non-compliance, identify barriers (cost, side effects, confusion about dosing), and trigger appropriate interventions—shipping alternatives to the patient’s pharmacy, initiating physician consultations, or connecting patients with financial assistance programs.

Chronic disease management could automatically adjust care plans based on biometric data, schedule follow-up appointments when indicators trend poorly, and coordinate care across multiple specialists without requiring patients to navigate the healthcare system themselves.

Emergency department diversion could detect patients likely to use the ED for non-emergency care, automatically connect them with appropriate alternatives (urgent care, telehealth, primary care appointments), and arrange transportation or virtual visits.

What This Means for Healthcare AI Builders

Siftwell’s agentic AI roadmap offers several lessons for healthcare technology founders:

Prediction is table stakes. Competitive advantage lies in what happens after prediction.

Operator expertise becomes more valuable, not less. Automating workflows requires deep understanding of how those workflows actually function in practice.

Integration complexity is the moat. Agentic AI requires connecting prediction systems, care planning tools, scheduling platforms, logistics providers, and communication systems. This integration complexity creates defensibility.

Start with non-clinical workflows. Healthcare remains cautious about automation in clinical decision-making. Preventive care coordination offers a lower-risk entry point for agentic AI.

Move incrementally from insight to action. Siftwell started with predictions, added contextual explanation, built care planning capabilities, and is now adding autonomous execution. Each step proved value before expanding scope.

The Ten-Year Vision Compressed into Weeks

Trey’s observation that “every week is like 10 years in the world of artificial intelligence” captures both the opportunity and challenge for healthcare AI builders. The technology is advancing faster than healthcare organizations can typically adopt it.

But for companies like Siftwell that combine operator expertise with technical capabilities, this creates an opportunity to build genuinely transformative systems. Not just better predictions, but AI that actually closes the loop from insight to action.

The future Trey describes—where AI automatically coordinates care based on individual patient context—isn’t science fiction. It’s the natural evolution of healthcare AI once you stop accepting that predictions are the end goal.

For health plans drowning in data but struggling to act on it, agentic AI offers a path forward. For healthcare AI builders, it represents the next competitive battleground. The question isn’t whether agentic AI will transform healthcare operations. It’s which companies will build it first and do it right.