Back to projects

2026 · Clinical AI · Full-stack engineering

Medivox

Building clinical AI workflows that make evidence, generation progress, and safety signals clear to the people using them.

Medivox is a healthcare technology platform. My contributions focused on the assessment-generation experience across mobile, backend, and AI-processing systems: helping clinicians move from source evidence to useful drafts with clear system feedback and practical safeguards.

Safe assessment generation

I helped build the systems that turn clinical evidence into assessment drafts while keeping evidence readiness and verification visible throughout the workflow. The work spans API contracts, background processing, and the clinician-facing product experience.

Trust and transparency in the interface

I designed and implemented feedback that explains what the system is doing: evidence status, generation phases, progress, retries, and meaningful confirmations. The goal was to make an AI-assisted workflow understandable rather than opaque.

Evidence safeguards

I contributed safeguards around clinical-document handling, including safety-status reporting and protections against untrusted document content. These controls support a workflow where the provenance and readiness of source material matter.

A note on confidentiality

This overview intentionally omits private source code, patient information, internal metrics, and implementation details. It is a public summary of my contributions, not a product specification.