Abnormal Security is a San Francisco-based cybersecurity company that uses behavioral AI to detect and block sophisticated email attacks including business email compromise, account takeovers, and supply chain fraud that evade traditional signature-based security tools. Their engineering challenges include building behavioral AI models that establish baseline communication patterns for every employee and flag anomalous messages, applying natural language processing to understand email intent and detect social engineering tactics, performing real-time classification at scale across millions of messages per day with extremely low false-positive rates, and continuously adapting detection models as threat actors evolve their techniques. The company's approach requires ingesting and analyzing rich contextual signals from email platforms, identity providers, and collaboration tools to build a comprehensive understanding of normal organizational communication. Abnormal Security's hiring patterns indicate demand for applied ML engineers, NLP specialists, security researchers, and distributed systems engineers building the real-time classification infrastructure that protects enterprise email.
| Location | Listings |
|---|