Machine Learning Engineer, Assistant Quality

Другое

• You will work on applied problems across agent quality, evaluation, personalization, retrieval, and orchestration. The ideal person is excited by shipping production systems, not pure research, and wants to help shape how Company’s assistant gets better over time through stronger signals, tighter feedback loops, and better end-to-end execution quality.
• Build and improve ML and LLM-powered systems that raise the quality of Company’s AI Assistant and autonomous agents across real user workflows.
• Design evaluation, benchmarking, and monitoring loops to measure assistant quality, model quality, and end-to-end system performance.
• Develop and iterate on signals, prompts, workflows, and model-driven logic that improve reasoning, planning, personalization, and task completion quality.
• Work across areas such as RAG, semantic search, recommendation-style systems, post-training or reinforcement learning, and agent orchestration where they materially improve product outcomes.
• Partner closely with product, design, and engineering teammates to understand customer pain points and ship high-quality production systems quickly.
• Contribute to the data and ML infrastructure needed to support robust experimentation, offline and online evaluation, and continuous model improvement.
• 2+ years of industry experience in machine learning, applied AI, or software engineering with significant ML ownership.
• Strong hands-on coding ability and a track record of shipping production systems, not just prototypes or research projects.
• Experience in one or more of the following areas: LLM applications, NLP, search, retrieval, recommendations, evaluation frameworks, agent systems, or personalization.
• Comfort working across both modeling and product engineering details, including experimentation, quality measurement, and production iteration.
• Proficiency in common ML tooling and strong software engineering fundamentals in languages such as Python, Go, Java, or C++.
• A pragmatic, product-minded approach. You know when to use sophisticated ML techniques and when simple, reliable systems are the better answer.
• A proactive, low-ego working style and excitement about learning quickly in a high-velocity environment.
• This role is hybrid (4 days a week in our San Francisco office)