Scott Campit
Agentic AI

Social Engineering Voicebots

Testing human vulnerabilities before real attackers do

Retailers know their employees are the weakest link in security, but traditional phishing tests only cover email. Real attackers use the phone — and most companies have no way to test how their teams respond to voice-based social engineering.

We built an agentic AI system that carries on natural phone conversations designed to probe employee awareness, complete with lifelike avatars and custom scenarios tailored to each client’s risk profile.

  1. Live phone-based social engineering simulation with natural conversational AI

  2. Vulnerability scoring model grading retailer exposure based on employee responses

  3. Web application with lifelike AI avatar synced in real time with the voicebot

  4. Custom caller identities and scenarios tailored per client — from IT impersonation to vendor fraud

  • Voice AI
  • Agentic Systems
  • Security Testing
  • Real-Time Avatars
  • Behavioral Scoring

Thinking about a similar problem?

I’d enjoy comparing notes. Tell me what you’re working on.

Email me scottcampit@gmail.com