JOB-1515
A seed-stage cybersecurity company is growing a compact team of quantitative researchers and engineers, and it needs a researcher to build machine learning models that spot account takeovers as they happen. Detection and forensics sit at the center of the product, so the researcher will own meaningful modeling goals and answer for the results.
What you’ll own
- Create new models and sharpen existing ones that flag compromised accounts, with accuracy targets that are deliberately extreme
- Untangle forensic questions by telling attacker behavior apart from genuine user behavior across a full activity history
- Run the whole research loop yourself: gather the data, scrub it, craft features and fit the model
- Pick architectures with live latency limits in view, then release and maintain the code you write
- Work in a compact, high-caliber team with clear individual accountability
Who thrives here
- Four or more years as a quantitative researcher, ideally at a leading trading firm or hedge fund
- Solid foundations in statistics, linear algebra, anomaly detection and behavioral modeling; NLP is a plus
- Experience taking models from raw data to production, including data cleaning, feature work and latency constraints
- Programming ability in Python, C++ or Rust, plus Git
- Comfort with noisy alternative data; mid-frequency equities experience is a bonus
- A bachelor’s degree or higher in a quantitative field; advanced degrees welcome
- Existing US work authorization is expected
On-site in the San Francisco Bay Area. Visa transfers considered, and sponsorship may be possible for exceptional candidates.
