Machine Learning Engineer, Signal

Role overview

This is a senior machine learning engineering role focused on the models that triage a high-volume global data stream in real time. The work centers on building and owning production systems that judge urgency, authenticity, and data quality so downstream consumers (trading desks and training pipelines) can rely on the output. It is an end-to-end position: framing the underlying problem, working with messy real-world data, prototyping, validating, and keeping models healthy after launch.

Responsibilities

- Take full ownership of urgency, quality, and information-extraction models across data curation, training, evaluation, and production deployment. - Transform noisy, multi-source raw inputs into clean, structured, and well-evaluated signal. - Design and maintain evaluation harnesses that hold quality steady as sources and conditions shift. - Optimize models to run inline on a live stream, balancing accuracy with low-latency serving. - Partner with data engineering to ensure models integrate cleanly into the surrounding pipeline. - Monitor production behavior, detect drift and regressions, and ship changes safely.

Requirements

- 5+ years of experience building and shipping production ML systems with direct ownership of models that served live traffic. - Applied background in one or more of: natural language processing, information extraction, time-series modeling, or large-scale ranking. - Deep expertise in Python and the modern ML stack (e.g., PyTorch, Hugging Face, scikit-learn), with the ability to work below framework abstractions when needed. - Demonstrated experience serving models at low latency and a habit of measuring what gets shipped. - Comfort with large-scale data tooling such as Spark, Ray, or Kafka to support training and inference workloads. - Comfortable owning open-ended production problems on a small team with broad technical scope.