Ai/Ml Product Manager
Columbus, Oh, Us
Leading a cross-functional team of 6 to develop a new hacker detection product that reduces detection time from hours to seconds and false positives from 100s to 1s per day from an idea to paying customers. Cyber attacks are on the rise. Stealing or ransomwaring data only needs one privileged account. Windows Active Directory (AD) is a store of privileged accounts for most enterprises and is a major target for hackers who breached the network perimeter. Many companies struggle to see AD attacks and have a significantly lower chance of stopping attackers. I work on a machine learning product, Attack Detector Solution (ADS), for detecting cyber attackers abusing AD and enabling a targeted response to prevent damage. Unlike AD monitoring solutions that focus on later stage attacks, ADS can find more attackers in less time because of its deeper coverage of techniques across the attack lifecycle. To make ADS successful, I contribute to: - market and competitor research, user and buyer research- market sizing and gross margin modeling- a product roadmap through problem discovery, prioritization and solution validation with business leaders and users- cloud architecture and data pipelines for continuous monitoring, detection and alerting- engineering of an extensible Bayesian detection system as an R package- design and implementation of a simulation lab for running and studying 40+ cyber attacks- design and implementation of a product monitoring framework- R&D of 30+ cyber attack statistical detectors (password spraying, kerberoasting, asp-roasting, ntds replication and dumping, AD recon, AD ACL abuse, etc.)- design of a digital marketing campaign- product presentations to stakeholders at prospective clientsTech stack: R, python, ELK stack, postgres SQL, HTTP, aws, docker, bash, git, Jenkins, jira, confluence, Windows logging-forwarding-collection system.
