Cursor
Software Engineer, ML Data Systems
About this role
Build and maintain the data infrastructure that powers Cursor's ML-driven code automation product. You'll own the full lifecycle of data pipelines—from fixing immediate bottlenecks to redesigning systems at scale—while ensuring correctness, cost-efficiency, and privacy guarantees.
What you'll do
- Design and ship data pipeline replacements while maintaining system reliability during transitions
- Define instrumentation requirements for new product surfaces and integrate telemetry end-to-end
- Debug performance issues across client, streaming, storage, and compute layers
- Establish data contracts and schema evolution strategies to prevent silent failures across consumers
- Optimize storage costs through retention policies, compression, and strategic data deletion
- Identify and close instrumentation gaps that impact model evaluation and experimentation
What they're looking for
- Apache Spark (Databricks or open-source)
- Large-scale data pipeline design and ownership
- Data modeling and schema design
- Production debugging across infrastructure layers
- Ray Data
- ClickHouse (preferred)
- Orchestration tools (dbt, Dagster)
- Privacy-aware data systems
Benefits
- In-person offices in North Beach (San Francisco) and Manhattan (New York)
- Flat organizational structure with small, talented team
- Ownership of high-impact systems that directly enable product teams
- Daily shipping culture with focus on quality and correctness
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Cursor
Cursor builds an AI-driven code editor used by millions of developers to transform how software is built. The company is hiring for infrastructure engineers, ML systems specialists, enterprise platform builders, security engineers, and customer success roles focused on driving adoption within large organizations.
- Website
- cursor.com
Likely interview questions
- Tell us about a large data pipeline you owned in production. What went wrong, and how did you decide whether to patch it or rebuild it?
- Describe your experience with Spark (Databricks or open-source). How have you optimized performance or debugged issues across compute, storage, and networking layers?