Scientific foundation for DLP (Data Loss Prevention) systems: real-time algorithms for confidential data flow detection, classification and blocking
From academic research to a working DLP platform: methodology, algorithms, validation results, and publications
Design Science Research Methodology (DSRM) — a rigorous 6-step approach combining academic rigor with industry practice.
Analysis of confidential data exfiltration in enterprise networks and quantification of business impact
Modeling insider threat, malicious leak, and accidental disclosure scenarios; defining target metrics
Designing the classification engine, content inspection, and ML-based sensitive data detection
Endpoint agent + network sensor + management console. Integrations with SIEM and Active Directory
Real enterprise pilots: accuracy, throughput, false-positive rate, user-impact measurements
Publishing in IEEE/ACM journals, filing patents, aligning with industry compliance standards
Deep Packet Inspection (DPI) of HTTP/HTTPS, SMTP, FTP, IM and other protocols. TLS inspection, SNI parsing, encrypted-traffic fingerprinting
ML-based engine automatically identifies PII, PCI, PHI, intellectual property and internal documents. Hybrid RegEx + ML, OCR for images/PDFs, EDM/IDM fingerprinting
Detected data flows are evaluated against policies in real time, with automatic response: block, quarantine, alert, audit log. Risk scoring + adaptive thresholds
IEEE Transactions on Network Security
This paper presents novel algorithms for real-time monitoring of confidential data movement in computer networks...
ACM Computing Surveys
This survey paper reviews machine learning approaches for confidential data classification...
Computer Networks
This paper presents methods for network protocol analysis in security applications...
Academic rigor, validated in industry pilots