Joshua Neil is a Co-Founder at Alpha Level, specializing in precision threat detection. With more than 25 years of experience working on data-driven solutions to USG and Industry enterprise security problems, he brings deep expertise in statistical methods for attack detection.
Joshua is a patented inventor and R&D 100 Award winner, known for creating the PathScan network anomaly detection system at Los Alamos National Laboratory. He believes statistical methods are under-utilized for attack detection and focuses on identifying attack-consistent events that are rare with respect to historic, predictive models.
Previously, he served as Chief Data Scientist at Securonix and Principal Data Science Manager at Microsoft for Microsoft 365 Defender, where he founded and ran the XDR Data Science team and developed groundbreaking ransomware kill chain detection. At Ernst & Young, he commercialized data-driven methods into production systems resulting in approximately $30 Million in revenue over 3 years.
He holds multiple patents in network anomaly detection and has authored numerous academic publications on authentication graphs, subgraph change detection, and cyber security.
- Alpha Level — Co-Founder: August 2023 – Present
- NeST Programme — ISOC Review Board Member: September 2022 – Present
- University of Washington — Industrial Review Board Member, Masters of Data Science: August 2020 – Present
- Securonix — Chief Data Scientist: June 2021 – August 2023
- Microsoft — Principal Data Science Manager, Microsoft 365 Defender: January 2018 – June 2021
- Ernst & Young — Senior Manager, EY Security Data Science: October 2014 – December 2017
- LANL (Los Alamos National Laboratory) — Research Statistician, Advanced Computing Solutions: September 2000 – September 2014
- Machine Learning
- Computer Security
- Statistics
- Honors-Awards: R&D 100 Award
- Patents: Non-harmful insertion of data mimicking computer network attacks, Path scanning for anomalous subgraphs detection, Anomaly detection for coordinated group attacks, Using new edges for anomaly detection
- Publications: Authentication Graphs, Towards improved detection of attackers in computer networks, Using new edges for anomaly detection
