Mithril

Mithril is building a cloud compute platform for AI developers offering high-performance GPUs, clean APIs, and infra primitives for training, fine-tuning, and serving models. Backed by Sequoia and Lightspeed. Hiring for multiple onsite roles in the SF Bay Area (Palo Alto or San Francisco) including General Software/Infrastructure Engineers (building batch + streaming workload engine, GPU scheduling, fault-tolerant execution, job DAGs), SRE Supply (managing GPU provisioning, spot bidding, node pool health across clouds and on-prem), Founding Product Manager (define roadmap for advanced ML infra users), and a general application for exceptional talent. Links to apply are provided on their Greenhouse job postings. They support distributed deep learning workloads, GPU autoscaling, long-running batch jobs, streaming inference, and related ML infra.