Context
Tolky is a ai-communications product shaped around a real user workflow. The work combines product framing, interface design, engineering, and production readiness in one delivery path.
tolky
Real-time browser speech translation for conversations that matter, without an app install.
Context
Tolky is a ai-communications product shaped around a real user workflow. The work combines product framing, interface design, engineering, and production readiness in one delivery path.
Challenge
The product needed to make a complex workflow feel clear while remaining reliable across devices, languages, and the operational constraints behind the experience.
Approach
We translated the core journey into small, testable slices, connected the right data boundaries, and kept security, accessibility, observability, and recovery visible from the first release.
Outcome
The result is a focused product foundation that can be measured in real use and extended without replacing the core architecture.
Delivery scope
Tolky captures speech in the browser, streams it over a live connection, and returns translated audio and transcript across Dutch, French, English, Turkish, German, and Spanish. The product is designed for clinics, municipalities, schools, and multilingual teams.
Discovery, UX direction, full-stack implementation, integration work, release support, and production verification. Stack: Next.js, Gemini Live, Web Audio API, WebSockets, GDPR, Realtime AI.