Back to all work
HealthTechWasfa Health·Jeddah, KSA·2024

An Arabic AI assistant for 200,000 patients

We added a clinically-guarded, Arabic-speaking AI assistant to a telehealth app, triaging symptoms and surfacing the right specialist while keeping doctors in the loop.

62%Triage automated
-40%Patient wait time
4.7/5Satisfaction score

The challenge

Wasfa's support team was overwhelmed by repetitive triage questions, and patients struggled to reach the right specialist quickly — especially in colloquial Arabic.

Our approach

We built a retrieval-augmented assistant grounded in Wasfa's vetted medical content, tuned for Saudi dialects, with strict guardrails, human escalation and a clinician review loop. Every answer is traceable to a source.

The outcome

The assistant now handles 62% of first-contact triage, cut average wait time by 40%, and lifted patient satisfaction to 4.7/5 — all under clinician oversight.

Let's build your next product

Tell us what you're building. We'll show you how an AI-native studio gets you there faster.