Talos: Scaling rare disease diagnosis with automated, iterative genomic reanalysis
90% diagnostic recovery. Microsoft's Talos cuts genomic review workload from thousands of variants to 1.3 per patient.
Why it matters
Microsoft Research shipped an open-source AI system that automates genomic variant analysis, reducing expert review burden by 99%+ while maintaining diagnostic accuracy—demonstrating real-world AI deployment in high-stakes medical workflows.
The key facts
10 to knowTalos recovered 90% of in-scope diagnoses
Reduced candidate variants to 1.3 per patient for expert review (vs. thousands in manual workflows)
Open-source system
Addresses bottleneck in genomic medicine: human expert review time
Published by Microsoft Research
90% in-scope diagnosis recovery rate
1.3 candidate variants per patient surfaced for expert review (vs. thousands without automation)
Open-source system (Talos)
Targets rare disease genomic medicine bottleneck
Focus on human-AI collaboration (automation + expert validation)
Go to the source
Microsoft Researchmicrosoft.com
Publisher excerpt: Talos was built to help resolve a major bottleneck in genomic medicine: human review time. The open-source system recovered 90% of in-scope diagnoses while surfacing just 1.3 candidate variants per patient for expert review.