AI in practice

AI embedded in real workflows—not added as a label.

My work sits at the intersection of product judgment, operational reality and applied AI: deciding where intelligence creates value, how humans remain involved, and how quality is measured.

01

Koi: AI at the point of care

Took an AI clinical assistant from idea to clinic workflows, connecting conversational diagnosis support, structured documentation and healthcare context.

02

Voice to structured clinical notes

Shaped real-time transcription experiences that converted conversations into useful clinical documentation while keeping doctors in control.

03

OCR for labs and prescriptions

Introduced document understanding for extracting and summarising lab and prescription information inside patient workflows.

04

AI-assisted licensing compliance

Led automation for Greek and collegiate design checks, reaching 89% and 91% accuracy and helping teams process tens of thousands of approvals.

05

Responsible productisation

Worked across prompts, evaluation, human review, adoption, workflow design and measurable accuracy—not just AI demonstrations.

89%Greek licensing accuracy
91%Collegiate licensing accuracy
3US clinics launched
FHIRInteroperable workflows