cDFN WatchTower Is Now Kervo AI
Same team, same platform, rebuilt as one system and given a name that matches what it does. Why we changed it, what Kervo AI is, and what it means if you run WatchTower today.

Today we're retiring the WatchTower name. cDFN WatchTower is now Kervo AI.
If you already run WatchTower, here's the short version. The team you work with is the same team. The platform is the one you use, rebuilt as one system. Nothing you rely on is going away. What has changed is how we think about what we're building, and the name finally matches it.
I want to explain why, in plain terms, because a rename from a security vendor deserves more than a new logo and a press line.
What we built WatchTower to fix
We started WatchTower at cloudDFN because we kept seeing the same thing in the security teams we worked with. Security had become a stack of separate purchases, each bought to solve whatever was loudest that quarter. Something for the attack surface. A vulnerability scanner. A service for leaked credentials. A cloud posture tool. A compliance platform for the auditors. A spreadsheet for vendors.
Every one of those tools did its job. None of them talked to each other. So a team would see thousands of findings and still miss the one thing that mattered: that an exposed server, an unpatched CVE, a leaked credential and an overpermissioned cloud role, each rated unremarkable on its own, added up to a route into their customer data.
WatchTower brought those jobs onto one platform. The part I'm proudest of is simple to describe: a vulnerability, the asset it sits on, a leaked credential and the compliance control it affects showed up as one connected picture instead of four tickets in four tools.
Why the old name stopped fitting
A watchtower watches. That's the promise in the name, and for a long time it described us well. We found things. We showed you what was exposed.
Two things changed.
The market filed us in the wrong drawer. Buyers heard "WatchTower" and understood attack surface management with some extra modules. That put us in bake-offs against tools whose whole identity is one slice of the problem, answering "why switch from the scanner I already have?" when the real question was "why run five tools that never agree?"
The product outgrew watching. Seeing everything is where security starts, not where it ends. Today the platform ranks vulnerabilities by severity, exploitability and business impact rather than a CVSS number. It investigates findings the way an analyst would, works out the fix and pushes it once a human approves. It keeps compliance evidence current every day and turns live posture into the report your board reads.
A name about watching was holding back a product about doing the work. So we changed it.
What Kervo AI is
Kervo AI is the AI control plane for security. In practice, that means a complete AI security team on one platform and one data model.
We organise it as a loop of five stages, because security isn't a set of tools. It's one job that never stops:
- Know: every asset and every exposure, outward from your domains and inward from the EDR, scanners and cloud accounts you already run, in one live inventory.
- Evaluate: every vulnerability reranked by severity, exploitability and business impact, and chained into the attack paths that lead somewhere worth protecting.
- Resolve: AI Analyst investigates each finding, works out its impact, finds the fix and pushes it once your team approves.
- Validate: controls checked continuously, with evidence collected automatically and mapped across every framework it satisfies.
- Orchestrate: AI CISO turns live posture into board-ready strategy, and the tools you already own feed one picture. Then the loop runs again.
The loop is the point. Each stage feeds the next, so the program runs continuously instead of in quarterly bursts before an audit.
A rebuild, not a relabel
It would have been easier to repaint WatchTower and call it a launch. We didn't. Kervo AI is the same platform rebuilt as one system, where every capability reads from and writes to the same data model. That's why a leaked credential, an unpatched CVE and an overpermissioned role become one attack path on their own, without anyone building an integration to find it.
I also know that "new security vendor" is a fair thing to be cautious about. So I'll say it directly: Kervo AI is not a first version. It's the platform our customers already trust, grown up and renamed.
Two ways teams use it
Lean teams, often one to ten people covering everything, run their whole security program in Kervo AI. The AI agents do the work they can't staff, on one contract.
Larger organisations keep the tools they've standardised on and connect them to Kervo AI. It resolves their conflicting asset records into one inventory and gives them the correlated picture no single product can produce.
Both run the same loop. The only difference is where the data comes from.
If you run WatchTower today
- Your team stays the same. Kervo AI is built and supported by the same people at cloudDFN LLP you already work with.
- Your platform is the one you know. Your environment, your findings and your integrations are where you left them.
- Your login moves. Sign in at platform.kervo.ai from now on.
- New capabilities arrive under the Kervo AI name, starting with AI Analyst and AI CISO.
If anything is unclear, write to us at support@clouddfn.com and you'll hear back from a person who knows your setup.
What the name promises
A name is a promise about what something does. WatchTower promised to watch. Kervo AI promises to do the work: find what's exposed, decide what matters, fix it, prove it, and report it, as one system that never stops running.
To everyone who ran WatchTower, told us where it fell short and pushed us to build what came next: thank you. Kervo AI exists because of you.
If you'd like to see it on your own environment, request a demo and we'll run it against your real infrastructure, not a demo tenant.
Want to see this on your own environment?
Your real attack paths, coverage gaps and compliance posture, rather than a scripted demo.
