AI Infrastructure Engineer: Model Training & Pipelines Remote

Zekai; United States; Remote; USD 200,000 - 320,000; Full time

Benefits offered by this job

Remote-first culture

Job summary

Zekai is seeking a Member of Technical Staff for Model Training (infra & platform) to own distributed systems and platform work enabling researchers to train advanced models. The role spans GPU infrastructure, inference optimization, and agent execution frameworks, with critical architectural decisions that affect research velocity and model performance.

The team designs pipelines for supervised fine-tuning, RL, and on-policy distillation, focusing on scalable training, performance, and

Qualifications

• Strong CS fundamentals in distributed systems, concurrency, networking and performance engineering.
• Experience designing, building and operating scalable backend systems or distributed computing platforms.
• Proficiency in Python and at least one systems-oriented language (C++, Rust, Go).
• Hands-on experience with GPU-based ML infrastructure, distributed training, model inference, or large-scale compute orchestration.
• Strong understanding of system performance: compute, memory, communication and fault tolerance.
• Experience designing reliable APIs, services and infrastructure components for complex workflows.
• Demonstrated technical ownership and ability to lead cross-functional initiatives.

Responsibilities

• Architect, build, and scale shared infrastructure for supervised fine-tuning, RL, and OPD.
• Design distributed training and execution pipelines coordinating GPU workers and simulations.
• Develop high-performance infrastructure for training and inference, optimizing GPU utilization and memory management.
• Evolve the core architecture of the agent ecosystem, including runtimes and serving APIs.
• Build robust systems for experiment tracking, dataset and model versioning, and observability.

Skills

Distributed systems

Backend scalability

Python

Systems language

GPU infrastructure

Performance engineering

Tools

PyTorch Distributed

DeepSpeed

Ray

Kubernetes

Job description

Zekai is seeking a Member of Technical Staff for Model Training (infra & platform) to own distributed systems and platform work enabling researchers to train advanced models. The role spans GPU infrastructure, inference optimization, and agent execution frameworks, with critical architectural decisions that affect research velocity and model performance.

The team designs pipelines for supervised fine-tuning, RL, and on-policy distillation, focusing on scalable training, performance, and