K Health: Powering AI-Driven Healthcare with a Personalized Cross-Platform Experience
When K Health was founded in 2016, the company set out to transform primary care. In 2024, they underwent a transformation of their own. One that required a highly capable engineering partner to bring to fruition.
Amid steady growth and even empirical evidence that K Health’s AI Physician Mode outperforms human doctors, the company set out to evolve the platform from a single-instance patient app into a white-label virtual care platform that allows hospitals and health systems to deliver personalized, AI-powered care at scale.
With insufficient internal bandwidth to simultaneously build toward the future and maintain existing functionality, K Health needed a trusted partner to extend their engineering capabilities immediately
“We needed a team tomorrow or we weren’t going to get it done in time.” Zachary Siegel VP of Engineering, K Health
K Health turned to Foxbox for the support necessary to rebuild the apps from scratch, natively in iOS and Android, optimizing for cross platform functionality and adaptability via Kotlin Multiplatform. With a focus on AI functionality, automated testing and scalability, our team hit the ground running as extensions of the K Health team.
Making AI-Driven Healthcare Happen
At the heart of the K Health experience is an AI-powered engine for symptom checking, triage, and scheduling—used by both patients and providers. Delivering that experience at scale required solving a deceptively complex problem: making advanced infrastructure feel completely invisible.
K Health supports live video visits powered by multiple video providers, each with their own distinct technical constraints. Our challenge was to abstract those differences entirely, so patients never had to think about what technology was powering their visit.
To achieve this, we engineered a unified video layer that handled audio, video, device selection, and real-time state changes consistently across platforms, despite fundamentally different event models under the hood. In healthcare, even small inconsistencies, like a muted microphone UI that doesn’t reflect actual audio behavior, can introduce serious risk, requiring rigorous testing and careful synchronization.
In parallel, we designed the system to support multilingual and accessible care without increasing product complexity. Language localization is resolved server-side, allowing the app to dynamically render content in a patient’s preferred language without hardcoded translations or client updates. The result is a flexible, future-ready foundation that scales across partners, languages, and accessibility needs, all without ever compromising the patient’s experience.
Building with Kotlin Multiplatform
Adopting Kotlin Multiplatform allowed us to build native iOS and Android apps with a shared, secure business logic. This shift dramatically improved scalability by reducing duplicate effort across platforms.
With Kotlin Multiplatform, we delivered fully native UI layers without compromising performance, resulting in lightning-fast interactions, smooth animations, and rock-solid functionality.We engineered for the future: optimizing for low-bandwidth environments, supporting multiple languages, and building a custom QA automation framework to validate daily releases and minimize regression risk.
We engineered for the future: optimizing for low-bandwidth environments, supporting multiple languages, and building a custom QA automation framework to validate daily releases and minimize regression risk. Every decision was made to enhance performance and long-term resilience.
“We validated every day… it really helps enhance the testing experience.” Manuel Marinez Senior QA Engineer, Foxbox
Embedded & Empowered
Our approach goes beyond writing code. From day one, we operated with full product ownership—proactively flagging issues, suggesting solutions, and adapting to shifting specs and timelines. We worked side-by-side with K Health’s internal engineers, designers, and product leads, navigating complexity with confidence and clarity.
This close collaboration enabled K Health to roll out new branded apps for major US-based health systems to provide AI-powered virtual care solutions, all while maintaining a single codebase and unified release pipeline. Because the architecture is built for flexibility, new deployments don’t require rework. Localization, customization, and expansion are possible on the fly. If a clinic wanted to launch in French tomorrow, they can do so without changing the code.
“They see something, they say something. They write it up, explore it, and suggest solutions.”Zachary Siegel VP of Engineering, K Health
Results That Scale
Today the K Health platform is powering virtual care programs at health systems that serve over 15 million patients. All without sacrificing development efficiency, code quality or the ability to adapt.
Together, our partnership doesn’t just extend capacity. It empowers K Health to move faster, think bigger, and scale smarter. And we’re proud to be along for the ride.
“The learning curve was steep, but now we can confidently build anything with Kotlin Multiplatform.” Antonio Arista, Senior Mobile Engineer, Foxbox
