Machine Learning Engineer (Evals and Voice Models)

At Aircall, a rapidly growing AI-powered customer communication platform, the engineering team focuses on developing innovative voice and messaging solutions. The role involves working on evaluation frameworks, voice models, and benchmarking systems to ensure high-quality AI agent performance across multiple channels. The team values collaboration, speed, and data-driven decision-making, with a global presence and a strong emphasis on AI innovation.

The candidate will design and implement evaluation systems for voice, chat, and messaging AI agents, including training and fine-tuning speech models and analyzing system design. They will also develop metrics, test pipelines, and monitoring tools to maintain quality and safety in production. The role requires experience in ML engineering, voice models, and automated evaluation systems, with a focus on reliability and continuous improvement.

This is a contract position based in San Francisco, requiring at least 8 years of experience in ML and 3+ years in ML engineering. The candidate should have expertise in supervised and unsupervised models, failure analysis, and voice model training. Strong communication skills are essential, and experience with synthetic data and LLM evaluation is preferred.