This role involves developing an iOS application for a real-time speech translation app that operates entirely on-device. The application will handle audio capture, speech recognition, translation, and text-to-speech, coordinated across multiple services on iPhones with hardware constraints. The developer will integrate vendor-provided ML models, manage background Over-The-Air (OTA) model delivery from AWS, build peer-to-peer device communication, and release the application through a federal accreditation pipeline. The responsibilities include owning the iOS application's architecture, performance, memory management, and production releases. This involves integrating on-device ML inference frameworks (Core ML, ONNX Runtime, WhisperKit), managing the real-time audio pipeline using AVAudioEngine, implementing voice activity detection and buffer management at hardware-thread priority, handling background model delivery with resumable multi-file downloads via URLSession, managing state persistence and network transition handling, and enabling peer-to-peer transcript sharing between devices using MultipeerConnectivity. The role requires utilizing Swift concurrency across all these features without introducing data races, deadlocks, or jetsam kills on 6GB devices.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
Job Type
Full-time
Career Level
Senior
Education Level
No Education Listed