Senior Machine Learning Engineer, Voice Agents

This is a fully remote position, open to applicants in France.

📋 Description

• Take charge of the architecture for significant components of the speech-to-speech open-source library, focusing on pipeline design, latency management, and real-time loop reliability.

• Implement new ASR, TTS, and end-to-end speech models while ensuring clean abstractions are preserved.

• Evaluate community pull requests, manage issues, release updates, and foster contributions from project collaborators.

• Develop the hf-voice developer API and streaming protocol, which encompasses session lifecycle, WebSockets/WebRTC transport, authentication, error semantics, and versioning.

• Create real-time GPU inference serving that includes concurrency, autoscaling, observability, and optimization of cost-per-session.

• Partner with Hub and inference teams to streamline the integration of voice agents into products and demonstrations.

• Transition hf-voice from a demonstration phase to production readiness through load testing, establishing SLOs, and ensuring graceful degradation.

• Produce documentation, examples, and templates for developers.

• Provide support for existing deployments, starting with the Reachy Mini fleet.

• Engage in public discussions about the work through blog posts, demonstrations, or conference presentations.

⛳️ Requirements

• Senior engineer capable of autonomously owning and advancing a significant portion of an architecture.

• Proven experience in building developer-facing infrastructure in an AI or developer-tools company, including inference APIs or agent infrastructure.

• Significant open-source contributions to a Python library.

• Proficient in async Python and distributed systems, including an understanding of their potential failure modes.

• Experience in deploying real-time systems involving streaming, WebSockets, WebRTC, audio or video pipelines, or live inference.

• Practical production experience with LLMs or multimodal models.

• Strong written communication skills and the ability to collaborate asynchronously and publicly.

• Passion for voice and conversational AI.

• Contributions to voice-agent frameworks such as speech-to-speech, Pipecat, LiveKit Agents, Vocode, or TEN.

• Contributions to llama.cpp or another low-level inference runtime.

• Hands-on experience with ASR, TTS, or end-to-end speech models, including evaluating latency and quality trade-offs.

• Experience with GPU serving, quantization, or on-device inference.

• Knowledge of audio pipelines, including VAD, echo cancellation, jitter buffers, barge-in, and turn detection.

• Experience delivering solutions for embedded or robotics targets.

• A public track record demonstrated through talks, blog posts, or demos.

🏝️ Benefits

• A workplace committed to diversity, equity, and inclusivity.

• An equal opportunity employer with a strong commitment to nondiscrimination.

• Reimbursement for relevant conferences, training, and educational opportunities.

• Flexible work hours.

• Options for remote work.

• Comprehensive health, dental, and vision benefits for employees and their dependents.

• Parental leave.

• Flexible paid time off.

• Opportunity for remote employees to visit offices in NYC and Paris.

• Workstation equipment provided as required.

• Equity participation for all employees.

• A community that actively supports the ML/AI community.