CareEZ — Clinical AI
Multimodal clinical AI for dysphagia / 照護食 standard adoption. Built on HKCSS 護食標準 + IDDSI.
Text, image, and voice classifiers that run live against the public API. Multilingual. Clinical-safety-biased. Governed by HKU + s.88 Full Linkage co-stewardship.
Live Demo
Try CareEZ Live
Live API · response in real-time · no PHI stored
Why CareEZ?
Multilingual Delivery
Clinical outputs in Cantonese, English, Tagalog, and Bahasa Indonesia — covering the full carer workforce in Hong Kong and across Asia.
Clinical Safety Bias
When in doubt, CareEZ over-refers. Conservative IDDSI classification prevents aspiration risk from being under-called in care settings.
Governance Inversion
Co-stewardship with HKU and s.88 Full Linkage means clinical standards and data governance are community-held, not commercially extracted.
Live API Endpoints
Three public endpoints. No auth required for evaluation. Responses in JSON.
IDDSI Text Classifier — describe a food, get the IDDSI level
curl -s -X POST https://seniordeli.com/api/iddsi-classify \
-H 'Content-Type: application/json' \
-d '{"description":"soft boiled rice"}'IDDSI Image Classifier — upload a food photo, AI reads the texture
curl -s -X POST https://seniordeli.com/api/iddsi-classify-image \ -F 'image=@meal.jpg'
Voice Aspiration Risk Screen — symptom checklist → risk tier + action
curl -s -X POST https://seniordeli.com/api/voice-aspiration-screen \
-H 'Content-Type: application/json' \
-d '{"symptoms":["coughing during meals"],"language":"en"}'Full API reference at seniordeli.com/api
- 100+
- Hospitals served
- 7
- Patents granted
- 17+
- Awards & recognitions
- IDDSI
- Compliant standard