Fullstack Engineer Data Platform
Posted Updated 2
views
Summary
Owns the data infrastructure pipeline between raw gameplay footage and model training/serving: managing GPU clusters and orchestration, building video preprocessing pipelines, optimizing disk/network I/O and production inference, and owning IaC, multi-region deployment, reliability, latency, and cost. Core stack includes Kubernetes, cloud infrastructure, and Python or Go.
You will own the infrastructure between raw gameplay footage and model training and serving. You will manage orchestration and GPU clusters, build preprocessing pipelines, optimize disk and network I/O and production inference, and own infrastructure-as-code, multi-region deployment, reliability, latency, and cost.
Responsibilities
- Own orchestration and GPU clusters, including scheduling, utilization, and capacity
- Build preprocessing pipelines for training-ready gameplay video data
- Analyze and optimize disk and network I/O bottlenecks
- Optimize production inference through batching, quantization, KV cache, and serving runtimes
- Own inference latency and cost metrics
- Own infrastructure as code, multi-region deployment, and reliability
- Make technical design and technology decisions and carry systems to production
Requirements
- Several years of deep infrastructure experience at a large technology company or serious lab, or five to six years working on hard infrastructure
- Current hands-on coding experience
- Experience owning a substantial system through production
- Knowledge of Kubernetes, GPU infrastructure, cloud providers, infrastructure as code, and multi-region deployment
- Experience with Python or Go
Benefits
- Meaningful equity
- Comprehensive medical coverage
- Dental coverage
- Vision coverage
- 401(k)
- Wellness and fitness perks
- Wellhub membership
- Mental health resources
- Paid parental leave
- Fertility benefits
- Maternal health benefits
- Generous PTO
- Daily meals at the NYC HQ
- Commuter benefits at the NYC HQ
- Learning and development stipend
