Agentic AI Engineer: Build Playable World Experiences

Astrocade; Palo Alto (CA); On-site; USD 180,000 - 280,000; Full time

Benefits offered by this job

Equity

Health insurance

Flexible PTO

AI Tools budget

Lifestyle Benefit

Fun Benefit

Parental leave

Childcare benefit

Workspace setup allowance

Commuter benefit

In-office meals

Retirement plan

Job summary

Astrocade in Palo Alto is seeking an ML Engineer on our AI team to own how our creation system behaves and where it fails. You’ll dig into creator sessions, build evals, and design the agentic systems behind the next version of creation on the platform, shipping to millions of players.

We are a ~20-engineer team collaborating across models, data pipelines, and infrastructure, working directly with our CTO and co-founders to define technical direction and product impact.

Qualifications

• 3–7+ years of ML engineering experience shipping models or AI systems.
• Strong CS fundamentals and clean, efficient Python code.
• Experience shipping AI systems to real users.
• Reasoning with probability and statistics; data-driven mindset.
• Builder with initiative and ownership of outcomes.
• Familiarity with Astrocade or interest in content creation platforms.

Responsibilities

• Own the quality of AI-generated content end to end, including model behavior, agent orchestration, prompting and failure modes.
• Build evaluation systems on production creator data to measure failure rates and quality.
• Design and build agentic ML systems that turn creator ideas into playable games.
• Debug complex issues across stack—from data to model to pipeline.
• Run fast experiments, measure rigor, and ship effective improvements.
• Help define how the AI team grows, including tooling and eval standards.

Skills

ML engineering

Shipping models

Python programming

CS fundamentals

Probability & statistics

Ownership mindset

Tools

Python

Job description

Astrocade in Palo Alto is seeking an ML Engineer on our AI team to own how our creation system behaves and where it fails. You’ll dig into creator sessions, build evals, and design the agentic systems behind the next version of creation on the platform, shipping to millions of players.

We are a ~20-engineer team collaborating across models, data pipelines, and infrastructure, working directly with our CTO and co-founders to define technical direction and product impact.