SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
DeepL is hiring a Senior Software Engineer to join the Voice team, which is building real-time voice translation and interpretation capabilities—the next frontier after DeepL's successful text translation platform. This is a 0-to-1 engineering role in a dynamic, cross-functional squad alongside researchers, product managers, and designers.
You will own end-to-end feature development across the full stack: building polished, modern UIs in React/TypeScript on the frontend and shipping backend services in Node.js/TypeScript (or equivalent) on the backend. The Voice team is tackling some of the hardest problems in real-time AI interaction: eliminating transcript instability ("flickering"), achieving sub-millisecond latency through chunked inference, and delivering natural-sounding audio streams.
Key responsibilities include shipping impactful features across the entire stack, debugging real-time edge cases (latency spikes, audio glitches, unstable networks), and collaborating closely with cross-functional partners to scope and solve hard product and technical challenges. You'll need a product-first mindset—understanding that 99% accuracy means nothing if audio arrives 3 seconds late—and high agency to identify and solve problems without waiting for permission.
The ideal candidate is fluent across the full stack, comfortable with real-time systems (WebRTC, WebSockets, streaming architectures), and leverages AI-powered tools (GitHub Copilot, Claude) to accelerate velocity. You thrive in ambiguity, prototype rapidly, and refactor for scale. DeepL operates a Unified Engineering Pipeline, so exceptional candidates may be matched to the highest-priority role across Core Product, Platform, Voice, or Enterprise based on skillset and company needs.
DeepL is a 1,000-person AI company founded in 2017, backed by Benchmark, IVP, and Index Ventures, serving 200,000+ business customers and millions of individuals across 228 markets. The company builds the entire AI stack in-house, from training LLMs on proprietary NVIDIA DGX SuperPODs to delivering real-time Language AI to 300+ million users.
