Keebler Health is a company that provides an AI-native risk adjustment platform for healthcare providers, accurately identifying chronic patient conditions and providing clinicians with insights and recommended next steps to keep patients healthy. Pulse 2.0 interviewed Keebler Health co-founder and CEO Isaac Park to learn more about the company.
Isaac Park’s Background
What is Isaac Park’s background? Park grew up fascinated by technology, learning to code in high school before earning a Bachelor of Science in Computer Science from Duke University. He began his career as a software developer building front-end frameworks before transitioning into product management, where he led diverse cross-functional teams through full product life cycles.
In 2009, Park co-founded Pathos Ethos, an innovation studio that helped startups and enterprise clients build mission-critical software in healthcare and defense. The company launched multi-million-dollar platforms and supported applications that reached over a million users.
After exiting Pathos Ethos in late 2022, Park joined the Duke Pratt School of Engineering as a faculty member at the Christensen Family Center for Innovation. In 2023, he co-founded Keebler Health, an AI-native risk adjustment platform for value-based care providers, and currently serves as the company’s CEO.
Formation Of The Company
How did the idea for the company come together? The idea for Keebler Health emerged from the co-founders’ shared passion for improving healthcare, particularly within Revenue Cycle Management (RCM). Isaac Park and Andrew Stickney connected through mutual friends and quickly discovered a shared interest in modernizing the back-end operations that burden healthcare providers.
Park’s interest was personal. After a challenging health episode, he became aware of the systemic inefficiencies in care documentation and reimbursement. Stickney, a seasoned operator and venture investor with two decades of experience, had seen firsthand how much time and money providers lost to manual and error-prone processes. And they joined forces with Kevin Hill, a Ph.D. neuroscientist and AI expert who had led data science at Redesign Health.”
Together, they launched Keebler Health to bring large language models (LLMs) to the frontlines of clinical documentation, diagnosis coding, and value-based care workflows.
Favorite Memory
What has been your favorite memory working for the company so far? Park said:
“One unforgettable moment was when a provider using Keebler identified multiple chronic conditions that had been missed by prior clinical reviews. Seeing our AI not just support clinicians, but actually surface overlooked risks that directly improved patient care, validated everything we set out to build.”
Core Products
What are the company’s core products and features? Park shared:
“Keebler Health is an AI-native risk adjustment platform built for healthcare providers operating in value-based care models. Our core offering uses large language models to process vast amounts of unstructured clinical data—handwritten notes, scanned charts, PDFs, visit summaries, and claims data—to identify chronic conditions that may have gone undetected.
Key features include:
— Prospective Suspecting: Surfacing potential diagnoses based on comprehensive chart review.
— LLM-Powered Ingestion: Parsing thousands of patient records per week at scale.
— Clinician Decision Support: Providing clinicians with explainable AI outputs and documentation flags for review.
— Compliance Reporting: Supporting accurate coding and audit readiness for risk adjustment and HCC submission.”
Challenges Faced
What challenges have Park and the team faced in building the company? Park acknowledged:
“Healthcare data is notoriously messy and unstructured. One of our biggest technical challenges was developing robust AI pipelines that could reliably parse handwritten notes, scanned documents, and fragmented health histories.”
“We’ve invested heavily in fine-tuning our models for healthcare-specific language and edge cases. We also built feedback loops with clinicians to ensure real-world relevance and clinical-grade accuracy. On the infrastructure side, cloud optimization efforts have helped reduce compute costs by over 50% year-over-year.”
Evolution Of The Company’s Technology
How has the company’s technology evolved since launching? Park noted:
“Our platform has evolved dramatically. Initially, we focused on structured EHR data and traditional RCM coding workflows. Today, we support full ingestion of scanned and handwritten documents—delivering accurate, prospective risk insights across diverse formats.”
“We’ve also improved our AI’s explainability, allowing clinicians to understand exactly why a condition was flagged. These advancements not only improve accuracy but also increase trust in our tools. Behind the scenes, we’ve seen a 3–5x performance improvement in throughput and up to 50% reduction in inference costs due to model optimization.”
Significant Milestones
What have been some of the company’s most significant milestones? Park cited:
— Customer Growth: In 2024, we grew from product launch to dozens of provider partners and a meaningful ARR run rate (exact figures confidential).
— Funding: Closed a $6 million oversubscribed seed round from top-tier investors, bringing total funding to $7.8 million
— Clinical Wins: Our AI surfaced conditions that were validated by clinicians but had previously been missed—proving our model’s real-world impact and ROI for physician groups.
— Pilot Success: Several pilots transitioned into full-scale rollouts after providers saw measurable ROI within weeks.
Customer Success Stories
When asking Park about the company’s customer success stories, he highlighted:
— A provider group with 7,000 patients used Keebler to conduct risk adjustment on their entire panel—something that would have taken months manually.
— Another group ran a pilot on 44 complex patients and uncovered 98 previously undocumented conditions, leading to significantly improved risk scores.
— One customer saw a 75x ROI in their first year by optimizing documentation and reducing reliance on manual chart reviews.
— On average, Keebler customers identify 15–30% more disease burden than traditional methods.
Funding
When asking Park about the company’s funding details, he revealed:
“Keebler Health has raised $7.8 million in seed funding from top-tier investors, including Freestyle, MBX Capital, New Stack Ventures, Underdog Labs, Ludlow Ventures, Everywhere VC, Primordial, the Tweener Fund, Hustle Fund, and several notable angels.”
“While we don’t publicly disclose exact revenue figures, we’re proud to say that our growth and efficiency metrics benchmark among the best in class for venture-backed software companies. We’ve been fortunate to see strong early commercial traction, driven by real customer value and a product that’s solving urgent problems in value-based care. It’s early days, but the response from the market has been both validating and energizing.”
Total Addressable Market
What total addressable market (TAM) size is the company pursuing? Park assessed:
“Risk adjustment alone is a $10–15B market in the U.S., and it’s growing quickly as more providers shift to value-based care. CMS aims for 100% of Medicare beneficiaries to be under value-based models by 2030—up from roughly 50% today.”
“However, the broader opportunity is even larger. Keebler sits at the intersection of risk adjustment, population health, quality measurement, care coordination, and patient engagement—a combined TAM estimated at over $80 billion to $100 billion. Our platform’s ability to extract insight from unstructured health data positions us to scale across these categories.”
Differentiation From The Competition
What differentiates the company from its competition? Park affirmed:
“Legacy vendors often rely on rules-based engines and partial sampling methods. Their solutions are labor-intensive, expensive, and often miss critical clinical context.
Keebler takes a radically different approach:
— LLM-Native: Our models are trained on the full spectrum of clinical data, including handwriting, scans, and longitudinal notes.
— Full Population Analysis: We don’t rely on sampling—we assess every patient’s full chart.
— Cost-Effective & Scalable: Our automation reduces costs by up to 70% compared to traditional services-based approaches.
— Explainability: Providers see not just what was flagged, but why—leading to greater clinician trust and adoption.”
Future Company Goals
What are some of the company’s future goals? Park emphasized:
“Our long-term goals are to:
— Reinvent population health through personalized, data-driven insights.
— Expand beyond risk adjustment into areas like predictive analytics, quality measurement, and care coordination.
— Improve outcomes at scale, enabling earlier interventions and better chronic disease management.
— Lower the cost of care by reducing administrative overhead and improving diagnostic accuracy.”
Additional Thoughts
Any other topics you would like to discuss? Park concluded:
“We’re at a turning point in healthcare, where AI can do more than automate—it can augment clinical thinking. Keebler Health is focused on building that future, where data becomes a superpower for clinicians, not a burden.
Thanks so much for the opportunity to share our story.”
