Modal

Seeking an ML Performance Engineer with 5+ years of experience to optimize ML systems for higher throughput and lower latency. The role involves working with inference engines like vLLM or TensorRT, understanding GPU architecture, and improving ML performance at scale.

Research role focused on building an evaluation framework for AI coding agents to improve developer experience on the Modal platform. This involves defining quantitative objectives, measuring performance, and translating insights into product improvements, while staying updated on agent advancements and customer use cases.

The role focuses on engineering for ML training systems, specifically optimizing the training process for production machine learning models, including language models. This involves working with frameworks like PyTorch and high-level training libraries, and optimizing for performance bottlenecks.

The Forward Deployed ML Engineer will partner with leading AI companies to help them achieve state-of-the-art performance on demanding AI workloads like LLM serving and model training. This role involves hands-on work with customers to architect and optimize their AI workloads on the Modal platform, contribute to open-source projects, and collaborate with product and sales teams. Requires 2+ years of ML engineering experience, familiarity with serving and training toolchains, and strong communication skills.

Backend engineer for an AI infrastructure company providing GPU access, instant container startups, and storage for training, batch jobs, and low-latency inference. The role involves building modern web applications end-to-end, working across the stack (TypeScript, Python, ClickHouse), and focusing on observability for large-scale AI workloads.

Customer Engineer role focused on AI/ML infrastructure, working directly with customers to debug and optimize workloads, while also shipping fixes, features, and automation to improve the core platform. Requires depth in either low-level infrastructure or ML/AI.