AI in healthcare is moving from pilots to production, but most HealthTech systems are not ready for it. In this webinar, Fedir Kompaniiets and Roman Burdiuzha of Gart Solutions explain what makes healthcare infrastructure harder than other industries, the layers an AI-ready platform needs, how HL7 and FHIR fit in, and the mistakes that stop projects from reaching production.
Published: 7 August 2025
Who it is for: Founders, CTOs and engineering leads building AI products for healthcare, from clinical documentation to diagnostics.
Key takeaways
- Data is the foundation: healthcare data is fragmented across legacy systems and formats, and it has to be cleaned and structured before any model is reliable.
- AI infrastructure is more than compute. It spans a data layer, a model layer, an integration layer and governance, monitoring and compliance, with compliance built in from day one.
- HL7 remains widely used for exchanging clinical data, while FHIR offers a modern, REST-style API that makes integration, validation and model training more efficient.
- Common mistakes include ignoring explainability and trust, poor data quality, weak security and launching without live monitoring.
- Real-time clinical use needs low-latency processing, audit trails and human-in-the-loop validation, not just a deployed model.
- For EU startups building healthcare AI, an EU cloud provider is worth considering from the start because of data rules.
Timestamps
- 0:03 Introduction
- 2:12 AI trends in healthcare
- 5:32 Why demand for healthcare AI is growing
- 8:48 What works today and what is still aspirational
- 13:09 What makes HealthTech infrastructure complex
- 19:42 The layers of AI-ready infrastructure
- 24:58 What infrastructure must support in clinical settings
- 26:08 The most common mistakes
- 32:43 HL7 and FHIR interoperability
- 38:08 How to start building an AI-ready HealthTech product
- 41:23 EU sovereignty and infrastructure choices
- 44:39 Q&A: HealthTech vs FinTech vs e-commerce, public cloud security
Speakers
- Fedir Kompaniiets, CEO and Co-founder, Gart Solutions
- Roman Burdiuzha, CTO and Co-founder, Gart Solutions
Related resources
- Building AI-ready infrastructure for HealthTech
- Practical AI applications in healthcare
- EHR integration
- Healthcare IT consulting
Want to discuss how this applies to your infrastructure?
Talk to our team, or browse all Gart Solutions webinars.
