Mindgard maps how attackers break AI. I map where your buyers are and build the programs that reach them first. Fifteen years of demand generation inside security companies, most of it selling to the AppSec teams now handed AI risk.
Mindgard profiles AI systems the way attackers do, mapping models, agents and tools before the first attack runs. Demand generation works the same way. Map the accounts, the people inside them and the places they gather, then build programs that reach them long before a form fills. Form fills are lagging indicators. Real demand is made upstream.
Everything here comes from mindgard.ai. No logins, no scraped data, no guesses about your numbers.
150+ publicly disclosed AI vulnerabilities, with names like Grok, Sora and Antigravity attached. Most security startups spend a year trying to manufacture a reason to be talked about. You publish one every few weeks.
The work left is conversion: each disclosure becomes a campaign, each campaign a set of named accounts, each account a conversation Sales is glad to take.
Mindgard was founded on research showing traditional AppSec could not address AI-specific risks. Somebody still has to own that risk, and in many enterprises it lands on the application security team.
I spent six years marketing to those teams at Veracode and three more at ThreatX. I know what they read, where they meet, and how skeptical they are of a vendor claim.
[un]prompted in Sydney and San Francisco, GrrCon, Offensive AI Con, CAMLIS. Five events in six weeks, and every one of them is a room full of people who build and break things.
The field play I rate highest for security buyers fits that calendar exactly: small practitioner dinners anchored to the conference, with your researchers at the table.
Your 2026 milestone is enterprise design partners. That shifts the motion toward named accounts, buying committees and intent, where a list of the right fifty companies is worth more than ten thousand leads.
I owned and deployed 6sense at ThreatX and ran signal-based orchestration for Sales at DeleteMe.
// Mindgard has no demand generation role posted today. So I scored myself against the standard Director of Demand Generation brief for a security startup, in its usual order, and flagged the row I only partly clear.
An engagement report with zero findings against the author would be the first thing I'd distrust.
Mindgard is Boston and London. My programs have been North American, run remotely across time zones.
What transfers is the engine itself. I'd build it in the US first, then lean on the London team for local context before spending a dollar there.
Application, API and runtime security, to the AppSec buyer, for most of my career. AI red teaming is new ground for me as a marketer.
So the first month goes inside the research: every disclosure, every talk, every question design partners ask. Nobody can market a finding they can't explain.
I've spent my career marketing to the people Mindgard now sells to. At Veracode it was application security, back when getting a developer to care about a scan was the whole job. At ThreatX it was API and application protection, with a CEO-hosted podcast that ran past a hundred episodes because practitioners kept listening. In between, at Capsule8, I was the first marketing hire and watched a runtime security company go from $300K to more than $3.5M in ARR.
What pulls me to Mindgard is that the research comes first. You already have the hardest ingredient in demand creation: a steady supply of things worth talking about, found by people who know how to find them. My job would be making sure every one of those findings reaches the accounts that should hear it, and comes back as pipeline.
I'd bring the plumbing too. I've built the marketing stack three times, owned 6sense, moved a team off ZoomInfo onto Apollo, and I model LTV and CAC so the budget conversation happens on numbers.
You don't have this role posted. I built this anyway, in an afternoon, because a page is a better sample of how I work than a PDF in an inbox.
Thirty minutes. I'll walk through how I'd turn the next disclosure into pipeline.