> Markdown version of [/jobs/ext/3298186-software-engineer-voice-platform](https://www.wearedevelopers.com/jobs/ext/3298186-software-engineer-voice-platform). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Software Engineer - Voice Platform - **Company:** Sela AI, Inc. - **Location:** San Francisco, CA, United States - **Experience:** Experienced - **Contract:** Permanent contract - **Skills:** Clean Code Principles, Software Debugging, Software Design Patterns, Distributed Systems, Python (Programming Language), Software Engineering, Data Streaming, Low Latency - **Published:** September 17, 2026 - **Apply:** https://www.indeed.com/viewjob?jk=6bcdf9a6f6ec47ff ## About the Role * 3-6 years of software engineering experience * Experience building and operating low-latency, real-time, streaming, or distributed systems in production * Strong Python or Go skills and a deep understanding of software design patterns and best practices * Strong performance-engineering instincts and experience using instrumentation to diagnose latency, reliability, or quality issues * Experience using AI coding tools or agents to ship maintainable, clean code * Experience in high-growth startups, ideally at Seed through Series B ## Description * Own major components of Sela's real-time voice platform, from architecture and implementation to reliability in production * Improve voice performance across latency, audio quality, turn-taking, interruptions, transcription, and speech generation * Build measurement, observability, and debugging tools that make complex voice issues easy to reproduce and diagnose * Integrate our voice stack with simulation and evaluation systems, including end-to-end voice-to-voice evals * Train, fine-tune, and deploy models for problems such as turn-taking, interruption handling, and voicemail detection * Evaluate new voice models and providers, and improve the scalability, reliability, and cost efficiency of a platform handling hundreds of thousands of calls per day