Company: Diadia Health
Co-Founders: Elena Ikonomovska (CEO) and Andrii Yasinetsky (CTO)
Website: https://www.diadiahealth.com/
About Diadia Health
Diadia Health is a diagnostic reasoning platform built for clinicians who want AI they can verify, not just trust. Every AI tool that touches patient care makes claims. Diadia is the platform built to prove them: every clinical claim it generates is sorted into one of three tiers, Supported by Science, Plausible, or Unsupported, so clinicians see exactly why the AI reached its conclusion, not only what it concluded.
Diadia was founded in 2023 by Elena Ikonomovska, PhD, and Andrii Yasinetsky to solve a problem the two of them saw across precision, functional, and longevity medicine. Elena spent over 20 years in applied AI and machine learning. Andrii spent his career building production AI systems and agent infrastructure, including roles at Uber and Google. Together they set out to give clinicians an AI tool built specifically for how they practice, one they could verify rather than simply trust.
Clinicians are drowning in multi-omic data, labs, metabolic panels, genetics, and gut health among others, with no way to read it all as one connected picture. Diadia analyzes that data as a single biological system rather than a set of disconnected reports, surfacing root cause relationships that would not appear if each test were read on its own. Every medical claim the platform generates is checked against a reasoning graph built on more than 310,000 peer-reviewed papers and over 700,000 genetic variants, then classified through that three-tier verdict system before it ever reaches a clinician.
The hardest problem Elena and Andrii had to solve was not building an AI that sounded confident. It was building one that could show its work. In an independent study, Diadia tested 3,035 medical claims across 36 patients against four leading AI models. The models agreed with each other no better than random chance on identical patient data, a cross-model Fleiss kappa of negative 0.045, the exact bottleneck Diadia was built to close. Diadia classified 67.1 percent of its own claims as Supported by Science, 30.2 percent as Plausible, and 2.7 percent as Unsupported, giving clinicians a transparent, auditable answer instead of a black box.
The platform now runs end-to-end around the patient visit, covering pre-visit diagnostic prep, in-visit multi-omic analysis, and post-visit branded patient materials, all trained on the clinical style of each individual clinician. Clinics using Diadia have cut chart review time for complex cases down substantially, based on Diadia internal modeling, freeing clinicians to see more patients without adding staff or sacrificing the personalization their patients pay for.
Since launch, Diadia has been recognized twice by TechRound alone, Top 50 Women in Startups and Tech and AITech45, plus wins from Femtech World and 100 Women in AI, building a track record of independent validation to match its own verification approach.
Diadia is not trying to replace clinical judgment. It is focused on making that judgment easier to trust, easier to scale, and easier to prove.
Want to find out what TechRound can do for your business?
