Software Engineer, ML Platform

Cursor is building the future of programming automation by creating the best development tool for professional coders. The company combines cutting-edge research, thoughtful design, and rigorous engineering in a lean, talent-focused organization that values truth-seeking, creative thinking, and shipping real code.

Cursor attracts engineers who thrive on ambitious technical challenges and want to work alongside researchers pushing the boundaries of AI-assisted development. The company operates with a flat structure that empowers individual contributors to own their projects end-to-end, creating an environment where your impact is both visible and meaningful.

As a Software Engineer on the ML Platform team, you'll build the infrastructure that transforms real product usage into training signals for continuously improving models. You'll work across one of four specialized areas: collecting and serving telemetry data from production use; creating shared environments and pipeline infrastructure for researchers; building observability and debugging tools; or shortening the path from research idea to validated runs on the GPU fleet. Your work will directly unblock researchers and product engineers, turning their recurring challenges into durable platform solutions that scale across the organization.

What you'll do

• Design and build core platform systems that ML researchers and product engineers depend on daily

• Collaborate directly with research teams to identify pain points and translate them into lasting infrastructure improvements

• Own the reliability, performance, and developer experience of the systems in your area of focus

• Ship incrementally in a high-ownership environment, measure the results, and iterate toward higher impact

• Operate production systems serving meaningful scale across data ingestion, pipelines, scheduling, or similar domains

What you'll bring

• Strong foundation in systems and infrastructure software engineering with proven experience building platforms others rely on

• Background operating production distributed systems at scale, whether event ingestion, data pipelines, job orchestration, or equivalent

• Comfort working across Linux, cloud platforms or bare metal infrastructure, and modern orchestration tools like Kubernetes, Ray, or similar

• Genuine interest in partnering with ML researchers and product engineers to understand their needs

• Ability to thrive in high-ownership environments where feedback cycles are tight and impact is direct

Nice to have

• Expertise in event ingestion, product analytics pipelines, or tracing frameworks like OpenTelemetry

• Experience with data frameworks such as Spark, Flink, or Ray for training data and ML dataset infrastructure

• Background building experiment tracking, evaluation tooling, or observability interfaces designed for researchers

• Familiarity with GPU scheduling, job queue systems, or compute cluster management

What they offer

• Full-time position based in San Francisco with in-person collaboration at offices in North Beach, Palo Alto, and Manhattan

• Work environment featuring well-stocked libraries and thoughtfully designed spaces built for focused, creative work