John C. Hale Email & Phone Number

United States

Irvine, California, Us

- Led successful cross-functional monthly forum for fraud prevention.- Developed a package for fraud strategy generation and threshold optimization.- Used XGBoost to build a deposit fraud model, as well as a new account fraud model. Deployed to production via a scheduled job in DataBricks.- Designed and built dynamic feature sets for the risk team using PySpark.- Built unsupervised models for bot detection using UMAP and HDBSCAN. - Explored using contextual multi-armed bandits for referral comms optimization.

Westlake, Texas, Us

Worked as part of a data science team within the fraud sciences & strategy organization.- Initiated and taught a Python learning bootcamp for internal fraud strategist team.- Led a monthly risk analytic forum.- Developed a login segmentation for risk scoring using K-Means and prototyped deploying it via a flask API. - Worked as part of a large cross functional team to build a solution for detecting victims of microcap pump-and-dump schemes.- Created an ensemble model consisting of logistic regression, XGBoost, and an MLP for detecting client impersonation in the call center.- Built an auto-regressive time series model for anomaly detection in applicant volumes.I also led a Python learning bootcamp and a monthly risk analytics forum.

Westlake, Texas, Us

Worked as part of a small distributed team to spread and implement DevOps practices across the organization. - Implemented server restart and service restart automations for 11 different applications using Operation Orchestration- Built two access provisioning automation and a generic automation for syncing distribution lists with databases using Operation Orchestration.- Designed an extensive deployment pipeline for an internal monolithic program, with the goal of getting to nightly builds using Bamboo.

Master Of Science - Ms, Data Science

Bachelor Of Science - Bs, Computer Science

Bachelor Of Science - Bs, Mathematics