About the Role:You'll work at the intersection of applied AI and large-scale systems, focusing on world generation pipelines, data-efficient training methods, and inference-time optimization for real-time generation. Your work will enable everyone make, play and monetize highly engaging games & 3D environments. This will include:
- Architect and implement game generation pipelines powered by LLMs and advanced agentic chains
- Optimize game iteration cycles, model accuracy, and generation speed across mobile and cloud stacks
- Scale the product
- Build feedback loops that learn from player & self-play actions to auto-test and improve games in real time
- Deliver context-aware creation via multi-turn dialogue and fine-grained parameter control
We're looking for engineers who have:- Built & scaled products with agents: Orchestrated multi-agent tool-calling with memory and planning, shipped to real users with live evals, tracing/observability.
- Ran LLM post-training: Curated clean preference data, automated data pipelines, and built eval harnesses.
- Built fast growing products: solo or with a friend launched products that grew to 1000s of users. Owned onboarding and growth loops and drove retention.
- Worked on GPU optimization & CUDA kernels: and similar.
Responsibilities:- Architect and implement game generation pipelines powered by LLMs and advanced agentic chains
- Optimize game iteration cycles, model accuracy, and generation speed across mobile and cloud stacks
- Scale the product
- Build feedback loops that learn from player & self-play actions to auto-test and improve games in real time
- Deliver context-aware creation via multi-turn dialogue and fine-grained parameter control
Strong candidates may also have:- Significant experience in software engineering and applied AI
- 1+ year full time experience at a hypergrowth startup
Compensation: - $150 an hour. Successful candidates might be offered a Full Time offer.
- Free housing, food, and transportation.
Growth OpportunityYou'll join a mission-driven team working at the frontier of AI world generation. We're HQ'd in SF and currently work all week around. In this role, you'll have the opportunity to:
- Own large parts of Roam's core generation and simulation stack spanning from multimodal data pipelines to agent orchestration and engagement modeling.
- Define new benchmarks for fun, retention, and interactive intelligence in 3D worlds using behavioral data across thousands of environments.
- Ship alongside a team that has already built products for 13M+ users, with 1000+ large creators ready to deploy our technology.
If you're excited to move fast, build at the edge of what's possible, and help define how AI inhabits 3D worlds, let's talk.
