Trotta - Trotta stops social engineering attacks pre-delivery
by•
Trotta uses proprietary machine learning to simulate attacker behavior and stop attacks before they reach humans or AI agents.
Replies
Best
Maker
📌
We built Trotta after working on AI penetration testing and attack prediction. We realized that identifying attacks and predicting them earlier still doesn’t stop people from being deceived. As models improve, social engineering will continue targeting humans and, increasingly, AI agents.
Our thesis became simple: if humans are the primary attack vector, why not remove them from the attack path?
That led us to pre-delivery interception: stopping malicious content before it reaches an inbox, conversation, AI agent, or tool. Our approach evolved from detecting suspicious messages to simulating attacker behavior and protecting the full action chain across email, messaging, APIs, and AI agents.
We’ve since filed five patent applications around this approach.
Trotta is the result: instead of asking humans or agents to recognize deception, we prevent them from ever having to make that judgment call.
We’d love your feedback!
Report
Cool thesis and product! Curious who your ICP is here. What types of companies are most susceptible to social engineering?
Report
Maker
@julian_alvarez Thank you very much Julian! We're focused on healthcare, aviation and financial providers - these folks are intentionally targeted because they've got suppliers/vendors/connectors/API providers with thousands and millions of employees. The cascading happens when one of the suppliers gets breached - 98% of these breaches starts with social engineering.
Report
A Slack and Teams integration would be huge since so much social engineering happens there now, and it would round out the messaging coverage nicely.
Report
Maker
@adnancwzc Thank you Adnan! Slack and Team integrations coming in hot ☄️this week!!!
Report
The approach of catching social-engineering payloads before they hit an AI agent feels genuinely thoughtful, especially since most security tools still assume a human is reading the message.
How different is this from an anti-virus or anti-malware software?
Report
Maker
@james_herdsen_limited Thanks James! Great question! Anti-virus only protects devices after malicious code exists. Closest analogy here is are actually firewalls - which based on specific rules and dynamic ones are focused on infra; the reality here is that the models have gotten really smarter, infrastructures are more secure than ever, but humans 🤔 not exactly. Trotta prevents humans and AI agents from being manipulated before any malware is ever delivered. We stop social engineering before action is taken.
Report
Really interesting approach. My biggest question is around false positives. If Trotta blocks or quarantines something legitimate, how do you balance security with business continuity, especially in environments where a missed email or API request could be costly?
Replies
Cool thesis and product! Curious who your ICP is here. What types of companies are most susceptible to social engineering?
@julian_alvarez Thank you very much Julian! We're focused on healthcare, aviation and financial providers - these folks are intentionally targeted because they've got suppliers/vendors/connectors/API providers with thousands and millions of employees. The cascading happens when one of the suppliers gets breached - 98% of these breaches starts with social engineering.
A Slack and Teams integration would be huge since so much social engineering happens there now, and it would round out the messaging coverage nicely.
@adnancwzc Thank you Adnan! Slack and Team integrations coming in hot ☄️this week!!!
The approach of catching social-engineering payloads before they hit an AI agent feels genuinely thoughtful, especially since most security tools still assume a human is reading the message.
@emirhana0qu Thank you very much Emirhan!
skillsync
Congrats on the launch Fela!!
@nars Thanks Nars!
How different is this from an anti-virus or anti-malware software?
@james_herdsen_limited Thanks James! Great question! Anti-virus only protects devices after malicious code exists. Closest analogy here is are actually firewalls - which based on specific rules and dynamic ones are focused on infra; the reality here is that the models have gotten really smarter, infrastructures are more secure than ever, but humans 🤔 not exactly. Trotta prevents humans and AI agents from being manipulated before any malware is ever delivered. We stop social engineering before action is taken.
Really interesting approach. My biggest question is around false positives. If Trotta blocks or quarantines something legitimate, how do you balance security with business continuity, especially in environments where a missed email or API request could be costly?