Software Engineer Model Lifecycle and Signals
Posted 6
views
Summary
Owns the path of frontier AI models from preview to production at Cursor: evaluating preview models, deploying and retiring them, building model-routing systems and experimentation/measurement platforms, and analyzing real-world data to improve model and product quality.
You will bring frontier models from preview to production and retire older models smoothly. You will evaluate preview models, build model-routing systems, create product signals and experimentation platforms, analyze real-world data, and turn evidence into improved models and products.
Responsibilities
- Deploy new frontier models to users and retire older models
- Evaluate internal and external preview models before launch
- Provide feedback to model-building teams
- Build model-routing systems that balance quality, speed, and cost
- Build product signals and experimentation platforms to measure model behavior and quality
Requirements
- Track record of shipping production systems
- Experience owning problems end to end
- Experience building and operating evaluation, measurement, or model-serving systems
- Experience with AI evaluation, experimentation, ranking, relevance, or routing systems
- Ability to turn quality questions into metrics, pipelines, and decisions
- Fluency with data and ability to collaborate with data scientists and researchers
- Knowledge of model and agent behavior
