ML Inference Engineer – Terminal-Bench Task Designer (vLLM / SGLang / TensorRT-LLM)

<p>Overview</p><p>We are building a benchmark that tests how well AI agents handle real expert work in a terminal. We are hiring ML inference and serving experts to turn real, paid tasks from their jobs into expert-level Terminal-Bench challenges. Each task includes the environment, a working solution, and an automated check that proves the result is correct.</p><p>***THIS IS A TASK BASED COMPENSATION MODEL. PLEASE CHECK THE APPLICATION QUESTIONS FOR MORE DETAILS.***</p><p>Responsibilities</p><p>• Design expert-level Terminal-Bench tasks that turn real, paid ML inference work from your own experience into challenges an AI agent must solve end to end.</p><p>• Define the files, systems and environments each task runs on, such as model serving with streaming, so the challenge is realistic, reproducible and command-line solvable.</p><p>• Write a working reference solution for every task and verify it runs end to end, sized for about 60 to 75 minutes of expert effort.</p><p>• Build automated checks that objectively prove a result is correct, including cases where ML outputs go wrong silently, such as numerical drift or metric bugs.</p><p>• Package tasks with Docker and Git so they are reproducible, and document the systems, tools, data and rules involved so reviewers can validate difficulty.</p><p>• Iterate on tasks based on reviewer feedback and agent test runs, adjusting difficulty and fixing ambiguities so each task stays expert-level, fair and objectively gradable.</p><p>Skills/Experience</p><p>• 3+ years doing ML inference or serving work hands-on (not only studying, selling or managing it), or a PhD plus active research in the subfield</p><p>• Did this work in the last 3 years and can name the systems, tools, data and rules used</p><p>• Trained, served or evaluated models at real scale (multi-GPU or production serving), not only notebooks or API calls</p><p>• Hands-on with PyTorch or JAX internals and serving stacks such as vLLM, SGLang or TensorRT-LLM</p><p>• Can explain how ML results go wrong without errors (numerical drift, metric bugs) and how to catch them</p><p>• Technical fluency in Python and the command line, and with files, CSV/JSON, APIs or databases</p><p>• Availability to start by early November 2026 and complete tasks of about 60 to 75 minutes each</p><p>Nice-to-have</p><p>• AI lab or LLM infrastructure experience</p><p>• Contributions to PyTorch, vLLM or SGLang</p><p>• GPU profiling experience</p><p>• Docker and Git proficiency</p><p>• Experience writing automated tests</p><p>Process</p><p>• Task vetting instead of an interview: complete a vetting task so the client can assess your task design</p><p>• Paid per task; each task takes about 60 to 75 minutes</p><p>• Work is delivered task by task, with reviewer feedback on submissions</p><p>• Expected start in early November 2026</p><br/><br/>Please mention the word **ASTONISHED** and tag RMmEwMTo0Zjg6MTQxOjQyMjk6OjI= when applying to show you read the job post completely (#RMmEwMTo0Zjg6MTQxOjQyMjk6OjI=). This is a beta feature to avoid spam applicants. Companies can search these words to find applicants that read this and see they're human.