Staff Engineer, Enterprise AI Systems

Ambral (YC S25); United States; On-site; USD 140,000 - 175,000; Full time

Benefits offered by this job

Equity ownership

Equinox membership

Free meals, coffee, and snacks

Health insurance

Unlimited PTO

Job summary

Ambral, a YC-backed startup, is building a replayable environment engine over real enterprise history to improve model performance. You’ll deploy into real workflows and work hands-on with customers to turn their data into scalable, evaluable environments.

You’ll work directly with the CTO, own production systems, and help design graders, environment factories, and evaluation sets. A strong track record in building production-grade software and handling large datasets is required.

Qualifications

• Experience building production-grade software.
• Ability to process large, messy datasets at scale.
• Strong intuition about systems and data structures.
• Able to translate fuzzy business objectives into tasks and signals.
• Comfort moving between research questions and production.

Responsibilities

• Build a replayable environment engine over real enterprise history.
• Collaborate with the CTO and customers to deploy into real workflows.
• Develop environment factory and graders to convert data into runnable environments.
• Mine useful tasks, trajectories, and evaluation cases from historical workflows.
• Design learning objectives and task definitions for evaluation.

Skills

Production-grade software

Large datasets

Systems design

Observability

Reproducibility

Production deployment

Job description

Ambral, a YC-backed startup, is building a replayable environment engine over real enterprise history to improve model performance. You’ll deploy into real workflows and work hands-on with customers to turn their data into scalable, evaluable environments.

You’ll work directly with the CTO, own production systems, and help design graders, environment factories, and evaluation sets. A strong track record in building production-grade software and handling large datasets is required.