
2-hour waits → 18-second language access
AI Interpreting
Jul 1, 2025
Nordhaven — a coastal city of 540 000 people — runs 14 neighborhood health-care centers, three urgent-care clinics and a 24/7 ambulance service. More than 30 % of residents speak a language other than Swedish at home, and the city routinely receives refugees whose first contact with public services is at a clinic reception desk.
The challenge
Before Tulka, staff booked freelance interpreters through a central dispatch office. Same-day slots were rare, travel time was unpredictable, and rare-language coverage (Tigrinya, Sorani, Uyghur) was almost nonexistent. Patients waited an average of 2 h 10 m, triage nurses juggled family members as ad-hoc interpreters, and medico-legal notes were often incomplete.
The solution
In March 2025 the Health Department deployed Tulka AI Interpreting on every exam-room tablet, front-desk PC and ambulance Toughpad. The roll-out took four weeks:
1. Pilot (Week 1) — Two high-volume clinics tested Tulka in Arabic, Somali and Russian.
2. Training sprint (Week 2–3) — 350 staff completed a 30-minute e-learning module plus on-site demos.
3. City-wide go-live (Week 4) — Geo-fenced QR codes redirected callers to Tulka’s browser client while the app badges were pushed to corporate iPhones.


— Dr Lena Isaksson, Chief Medical Officer
Results after 90 days
Wait time for language support fell from 130 min → 18 s (median).
Walk-in capacity rose 22 % with no extra clinical staff.
Rare-language coverage jumped from 18 to 57 languages available 24/7.
Patient-satisfaction on the “felt understood” metric reached 97 % (up from 61 %).
Cost savings: € 138 000 annualised, thanks to pay-per-minute billing and zero travel fees.
Compliance boost: Encrypted call logs now attach automatically to EHR entries, satisfying GDPR audit requirements.
Impact stories