Principal Engineer, Self-Improving AI Agents

NVIDIA Corporation; Santa Clara (CA); On-site; USD 224,000 - 431,000; Full time

Job summary

NVIDIA Corporation in Santa Clara, CA seeks a principal engineer to define evaluation strategy for agents and enterprise-scale systems. You will design how eval results guide self-improvement across skills, plugins, and harness configurations, with sandboxed tests and human sign-off.

Lead cross-team evaluation efforts, mine traces for failures and patterns, and ensure trustworthy improvements. Strong emphasis on experiment design and senior IC leadership.

Qualifications

• BS or MS in Computer Science, Engineering, or related field (or equivalent experience).
• 12+ years building large-scale systems, including evaluation systems for ML or agents and data pipelines.
• Hands-on experience building and evaluating agents on an open or closed agent harness.

Responsibilities

• Build evaluation that works across harnesses with automated checks for regressions.
• Mine production traces for failures, usage patterns, and performance hot spots.
• Turn data into eval cases and improvement signals.
• Develop self-improvement loop for evolving harnesses with sandboxed evals.
• Ensure only measurable gains ship, with human sign-off and limits on agent changes.
• Create realistic enterprise environments for safe long-running tasks.
• Measure multi-agent swarms and track permissions across hops.

Skills

Python

Go

Rust

Experiment design

Senior IC leadership

Education

BS or MS in Computer Science or related

Job description

NVIDIA Corporation in Santa Clara, CA seeks a principal engineer to define evaluation strategy for agents and enterprise-scale systems. You will design how eval results guide self-improvement across skills, plugins, and harness configurations, with sandboxed tests and human sign-off.

Lead cross-team evaluation efforts, mine traces for failures and patterns, and ensure trustworthy improvements. Strong emphasis on experiment design and senior IC leadership.