poolside

Build full-stack product surfaces that make Poolside’s model-training workflows visible, usable, and fast. You’ll own features end to end—from backend services and APIs to polished web interfaces—for researchers and engineers building AI models.

Build and operate the data and analytics foundations powering Poolside’s Model Factory and AI research workflows. You’ll partner directly with researchers to create reliable, scalable systems for experiment data and accelerate model development.

Lead Poolside’s multimodality effort, building image-input and eventually native multimodal capabilities for frontier coding models. You will establish a new team, set technical direction, and run end-to-end research and training at scale.

Build Poolside’s greenfield Experiment Platform, creating backend services, APIs, and a CLI that connect research workflows with training and evaluation infrastructure. This remote role sits at the intersection of research, infrastructure, and agent platform engineering.

Build the infrastructure that schedules GPU workloads and serves large-scale model inference for Poolside’s frontier AI research. This role focuses on improving GPU utilization, inference throughput, latency, reliability, and research velocity.

Own and evolve the build, CI/CD, and developer tooling infrastructure that enables Poolside’s researchers and engineers to build, test, and ship efficiently. This infrastructure role combines Bazel, Python ecosystem management, Kubernetes, AWS, and internal tooling with platform ownership and incident response.

Build and operate secure, reliable sandboxed execution environments for AI agents at massive scale. This platform engineering role combines distributed systems, containerization, virtualization, and performance optimization to support agent research and product development.

Join Poolside’s Applied Research team to turn pre-trained LLMs into highly capable, well-aligned systems for coding and software development. You’ll combine hands-on research with production engineering across post-training algorithms, data pipelines, and large-scale experiments.