Dear OpenAI Talent Acquisition team,
I would like to apply for the Machine Learning Engineer position you have open on your careers page. machine learning engineering has been my main line of work for several years, and I would be glad to contribute it to your team.
I spent some time on OpenAI before writing, and what caught me was your work on large-scale model serving and the engineering notes your team keeps posting on inference at scale. This looks like a strong time to join, and I would gladly put my machine learning engineering experience to work on it.
From the posting, the three areas you care about most are training and serving models at scale, model optimization and latency and production ML pipelines. Those decide whether a machine learning hire delivers, and I have solid results in each.
On training and serving models at scale, I work with PyTorch, TensorFlow and CUDA. As an Machine Learning Engineer at Waymo, I built a model-serving system that held p99 latency under 40ms at 10k requests per second. On top of that, I built the shared training pipeline the whole research team now runs on.
For model optimization and latency, I rely on ONNX, quantization and Triton. In my time as a Machine Learning Engineer at Waymo, I quantized a recommendation model and cut inference cost by 55% with no drop in accuracy.
On production ML pipelines, I bring Kubernetes, Ray and MLflow. Working as a Machine Learning Engineer at Waymo, I built an end-to-end training pipeline that took retraining from days down to hours. On top of that, I wrote the model-rollout playbook the whole team follows.
I would be glad to take you through any of this in an interview and lay out why I fit. I am ready to ship models into production, help the team move fast, and keep growing with it.
Thanks for reading, and I hope we can find time to talk.
Yours sincerely,
Theo Script
theo.script@gmail.com