Member of Technical Staff, Compute

Mirendil; San Francisco (CA); On-site; USD 300,000 - 400,000; Full time

Job summary

Mirendil Inc. is building a frontier AI research company and seeking an engineer to orchestrate a large GPU fleet across many providers. You will influence how much compute goes to research by designing fleet bring-up, scheduling, reliability, and capacity strategies.

The role focuses on turning GPUs into a seamless, productive compute resource for researchers, with a base salary and meaningful equity depending on experience.

Responsibilities

• Fleet bring-up and lifecycle - take new clusters from delivered to running real jobs fast, and keep them healthy as they age
• Scheduling and allocation - decide what runs where, so pretraining, RL, and inference share capacity without fighting over it
• Reliability at scale - detect and route around failing hardware automatically, before it costs us a training run
• Capacity strategy - plan with research what compute we'll need, and help decide what we buy, from whom, and when
• Utilization - measure what every GPU-hour produces, and keep raising it

Job description

Mirendil

Mirendil is a tech-first company focused on solving core bottlenecks that unlock step-change acceleration across science and technology. Our first goal is to democratize frontier AI R&D across scientific disciplines. We are building a frontier AI research company and training our own models end-to-end.

The Role

Our research moves only as fast as our compute. We're looking for an engineer who can turn a lot of GPUs across a lot of providers into one fleet that researchers never have to think about. Your work determines how much of our compute actually goes to research. Some example areas you might work on:

• Fleet bring-up and lifecycle - take new clusters from delivered to running real jobs fast, and keep them healthy as they age

• Scheduling and allocation - decide what runs where, so pretraining, RL, and inference share capacity without fighting over it

• Reliability at scale - detect and route around failing hardware automatically, before it costs us a training run

• Capacity strategy - plan with research what compute we'll need, and help decide what we buy, from whom, and when

• Utilization - measure what every GPU-hour produces, and keep raising it

We offer a base salary of $300,000-$400,000 and a meaningful equity grant, depending on experience and background, along with competitive benefits.