Canyon Bryson

Canyon Bryson

AI Engineer + Full-Stack Developer

I write reuseable, pattern-based code that is easy to understand and maintain.

7 years experienceComputational math + CS degreeLead Engineer at ORTHOATHLETEOrthoHCP shipped in 30 days
Download Resume
Back to Projects

OrthoHCP

Healthcare onboarding platform

Comprehensive onboarding for Orthopedic Surgeons, Physical Therapists, Trainers, and other HCPs with AI-powered video interviews and media portfolios.

FlutterNext.js 15React 19TypeScriptSupabaseOpenAI (Realtime)

Try the App

Available on iOS and Android

App StoreGoogle Play
2apps— Flutter + Next.js
8roles— Role-based access
5buckets— Media storage

Problem & Context

Ortho Marketing needed a scalable way to onboard healthcare professionals with rich profiles, interviews, and media portfolios. Existing workflows were manual and fragmented, making it hard to standardize quality and move candidates through approval.

OrthoHCP unifies onboarding into a structured, multi-step profile flow with AI interviews, storage-backed media uploads, and real-time updates across mobile and web.

Constraints

  • Multi-platform delivery: Flutter mobile + Next.js web
  • Secure access: RLS, JWT sessions, and signed URLs
  • Media-heavy workflows: video interviews, portfolios, and documents
  • Realtime UX: live updates and upload status tracking
  • Healthcare-grade onboarding: clear, guided profile creation

Solution Overview

  1. 1. Profile setup
    Multi-step onboarding for roles, specialties, and bio
  2. 2. Face alignment
    Guided positioning for consistent interviews
  3. 3. Interview playback
    AI-generated questions via TTS or prerecorded audio
  4. 4. Video recording
    High-quality capture with retries and progress tracking
  5. 5. Transcription
    Live speech-to-text during interviews
  6. 6. Upload & processing
    Secure storage, metadata tagging, and media organization

The platform guides HCPs through profile creation, runs AI interviews in real time, and organizes all media in Supabase storage for easy review and approval workflows.

Technical Highlights

  • Row Level Security — RLS policies ensure each user only accesses their own profiles, interviews, and media.
  • Media storage organization — Dedicated buckets for videos, images, documents, audios, and general media.
  • Realtime updates — Supabase Realtime keeps profile status and upload progress synced.

Key Decisions

Supabase as the backend backbone

Used Supabase for Postgres + RLS, authentication, real-time subscriptions, and storage. Edge Functions handle AI interview orchestration and secure server-side logic.

Tradeoff: Leans on platform features and conventions, but dramatically accelerates delivery and reduces infrastructure overhead.

WebRTC-powered AI interviews

Real-time OpenAI interviews run over WebRTC with live speech-to-text and text-to-speech. Interview state is coordinated through Riverpod providers on mobile.

Tradeoff: More complex client logic, but delivers a natural interview experience and higher quality responses.

Two-front-end strategy

Flutter powers the cross-platform app while Next.js provides the web dashboard. Shared Supabase auth keeps accounts consistent across devices.

Tradeoff: More surface area to maintain, but enables a best-in-class mobile experience and a polished web portal.