Built by Security Practitioners.
For Security Teams.
DefendML was built by people who have sat on the auditor's side of the table — and who understand what security teams actually need when they say "prove your AI is tested."
DefendML runs on a hybrid team — human expertise and AI agents working autonomously 24/7.
Sareth Dustin Sovan
Founder & Chief Architect
Enterprise Security Leader • SOC 2 Type II & ISO 27001 Programs • Palo Alto Networks Ecosystem
20+ years in IT and cybersecurity. Chief architect leading enterprise security systems for large-scale organizations. Led multiple SOC 2 Type II and ISO 27001 audit programs. Designs the scanning infrastructure and framework mapping logic that powers every DefendML evidence report. Deep specialization in adversarial machine learning, swarm AI attack simulation, and security framework alignment.
Jordan
Chief Revenue Officer
Handles outreach, qualification, and enterprise pipeline. Operates autonomously — identifies prospects, personalizes outreach, and advances deals 24/7 without downtime.
Sofia
Chief Marketing Officer
Manages LinkedIn content, SEO strategy, and threat intelligence publishing. Monitors competitor activity and publishes DefendML Labs findings around the clock.
Morgan
Founding Engineer
Built the 9-agent swarm from the ground up. Ships fixes at 1am. Deployed every worker version. Has never asked for a raise.
DefendML is AI-native by design. Our AI agents operate autonomously 24/7 — handling sales, marketing, and engineering alongside our human founder. This is what the future of work looks like.
Interested in working with us? Get in touch →How We Work
Four principles that drive every product and engineering decision at DefendML.
Attack First
We find vulnerabilities before adversaries do. Every product decision starts with the attacker's perspective.
Evidence Over Assertions
Claims that an AI system is safe are not enough. We produce structured, auditable evidence that security teams can stand behind.
Honest Positioning
We test AI applications pre-deployment. We don't provide runtime protection, and we don't overstate what a single scan can prove.
Practitioner-Built
DefendML is built by people who have been through SOC 2 audits, incident reviews, and security assessments — not just read about them.
Join Our Mission
We're building the offensive AI red team testing layer the security industry needs — starting with a library of 415 adversarial scenarios and evidence reports that hold up to enterprise scrutiny.
DefendML is a small, focused team. We work with security engineers and AI practitioners who want to build the offensive testing layer the AI security industry is missing. If that's you, reach out.