Senior Software Engineer | Voice | Full-stack

DeepL DeepL · AI Frontier · London, United Kingdom · Engineering

Senior Software Engineer for the DeepL Voice team, focusing on building real-time interpretation features from scratch. This role involves end-to-end feature development across the full stack, including frontend UIs, backend streaming services, and AI inference integration, with a strong emphasis on tackling real-time complexities and optimizing for user experience.

What you'd actually do

  1. A "0-to-1" Builder Mindset: You thrive in ambiguity. You don't need a roadmap to build; you need a problem. You are comfortable prototyping a new feature in the morning and refactoring it for scale in the afternoon.
  2. Product Obsession: You care about the _flow_ of the conversation. You understand that "99% accuracy" means nothing if the audio arrives 3 seconds late. You optimize for the human experience.
  3. Ship End-to-End Features: Build and deploy impactful features across the entire stack, from high-performance frontend UIs to streaming audio services on the backend and AI inference integration.
  4. Tackle Real-Time Complexity: Debug and resolve challenging edge cases unique to real-time communication, including sub-second latency spikes, transcript flicker, audio glitches, and unstable network environments.
  5. Collaborate Cross-Functionally: Partner closely in a tight-knit, cross-functional squad alongside PMs, Product Designers, and AI Researchers to scope, shape, and solve hard product and technical challenges.

Skills

Required

  • React
  • TypeScript
  • Node.js
  • WebRTC
  • WebSockets
  • streaming architectures
  • latency-sensitive systems

Nice to have

  • AI-powered tools (like GitHub Copilot or Claude)

What the JD emphasized

  • 0-to-1 engineering environment
  • real-time
  • sub-second latency spikes
  • transcript flicker
  • audio glitches
  • unstable network environments
  • building proprietary models
  • streaming architecture

Other signals

  • 0-to-1 engineering environment
  • building proprietary models
  • streaming architecture
  • real-time voice translation