The document summarizes a speaker verification system that uses voice biometrics to authenticate users. It describes the objectives, phases, features, and specifics of the system. The system extracts voice features from training samples to create feature matrices and thresholds for comparison during authentication. When a claimant speaks, features are extracted and compared to the claimed speaker's thresholds to determine if the voiceprint matches within 20% for verification. The presentation concludes with proposed enhancements to the system.