Senior ML Engineer

About the job

Responsibilities

• Own the machine learning for Dodo's agents, including models in the call path and evaluation processes.
• Enhance agent capabilities in understanding callers, selecting actions, and completing bookings across voice, text, and web.
• Develop evaluation pipelines and judges to measure agent quality, detect regressions, and validate model and prompt changes.
• Fine-tune, distill, and route models to meet real-time voice latency and cost constraints while improving quality.
• Construct data loops to convert conversation outcomes, practice feedback, and corrections into training signals.
• Manage speech and voice quality aspects related to model integration, including transcription, turn-taking, interruption handling, and latency.
• Collaborate with product engineers to deploy model improvements into production and assess their impact on bookings and customer outcomes.
• Define the strategic direction for Dodo's model utilization in a rapidly evolving AI landscape.

Requirements

• 3+ years of experience building and deploying machine learning or LLM systems in production.
• Extensive experience in LLM evaluation, fine-tuning, prompt and model routing, or speech systems.
• Strong software engineering fundamentals with experience writing and owning production code.
• Proficiency in measurement and designing evaluations for ML systems.
• Pragmatic approach to latency, cost, and reliability in real-time systems.
• Experience working with messy real-world conversation data and understanding healthcare privacy obligations.