AI Engineer

Introspection Inc.; San Francisco (CA); On-site; USD 150,000 - 230,000; Full time

Job summary

Introspection Inc. in San Francisco, CA is seeking an AI Engineer to architect and implement the systems behind long-horizon vertical agents and to own problems end-to-end.

You will work on agent recipes, evaluation frameworks, production evidence, and learning loops that turn observed failures into better releases, bridging cutting-edge research with production systems. Collaboration with customers and the research community is key to pushing the boundaries of what’s possible.

Qualifications

• Deep experience building LLM-powered systems.
• Experience with agentic workflows at scale.
• Experience bridging research and production.
• Ownership of problems end-to-end.

Responsibilities

• Architect and implement systems behind long-horizon vertical agents.
• Develop agent recipes, evaluation frameworks, production evidence, and learning loops.
• Collaborate with customers and the research community to push boundaries.

Skills

LLM systems

Agent workflows

AI safety

Evaluation frameworks

Prompt engineering

Infrastructure

Job description

We're looking for an AI Engineer who can architect and implement the systems behind long-horizon vertical agents. You'll work on agent recipes, evaluation frameworks, production evidence, and learning loops that turn observed failures into better releases. This role bridges cutting-edge research with production systems.

You might be a fit if...

You have deep experience building LLM-powered systems and agentic workflows at scale.

You're passionate about AI safety, evaluation, and systematic improvement of AI systems.

You can think across the stack, from prompt engineering to infrastructure, and can own problems end-to-end.

You're excited to collaborate with both customers and the research community to push the boundaries of what's possible.

About Introspection

Introspection is an applied AI team building the infrastructure for long-horizon vertical agents. We turn production conversations into evidence for better behavior, stronger quality gates, and more reliable releases.

We're a small, high-signal team based in San Francisco that values deep thinking, creative problem-solving, and shipping code that matters.