Technical Lead — Real-Time Voice Translation Platform

H2B GroupVoice translation

Praca w pełni zdalna, w europejskich strefach czasowych., europejskie strefy czasoweContractLead

Salary not disclosed

Apply NowOpens the employer's application page

Job Details

Have documented production experience with real-time audio processing systems, especially streaming voice pipelines.
Have practical experience in at least two of streaming Speech-to-Text (STT), streaming or incremental Machine Translation (MT), and streaming Text-to-Speech (TTS).
Have production WebRTC experience, including audio transmission under unstable connections and mobile network constraints.
Have experience instrumenting and measuring latency on end devices, analyzing median and p95 distributions, and interpreting outliers.
Be able to assess technology providers against compliance requirements, especially EU data hosting and zero-retention guarantees.
Have very good Python skills for backend development.
Know frameworks for orchestrating streaming audio pipelines, such as Pipecat or equivalent solutions.
Have Technical Lead or Architect-level experience, including independently making, documenting, and defending technical decisions.
Be able to analyze trade-offs between latency, quality, cost, scalability, and data privacy.
Have very good spoken and written English for client meetings and precise technical documentation.
Be ready to start at short notice and available full-time for the first two weeks of the project.

Design the solution architecture and compare STT → MT → TTS with end-to-end speech-to-speech models.
Design streaming audio processing that can begin playback before the source utterance is complete.
Evaluate STT, MT, TTS, and WebRTC audio transport providers for streaming and compliance requirements.
Define and measure latency budgets across pipeline stages, including sources of delay and variation.
Compare TTS options for voice naturalness, speaker similarity, streaming performance, language support, and cost.
Define testing criteria for transcription and translation quality and measure time-to-first-audio, percentiles, and outliers.
Set measurable pass/fail acceptance criteria for moving to the Alpha stage with the client.
Model audio-processing costs for different platform usage scenarios.
Assess which pipeline components could be moved onto mobile devices.
Set up the repository, CI/CD, application skeleton, and basic authentication; prepare architecture documentation and recommendations.

View Full Description & ApplyYou'll be redirected to the employer's site