01
We start with your threat model
Not everyone needs the same. Someone with visible wealth, an executive, a person with public exposure or a company handling client data face different risks than a private individual. Before touching anything we define what —and whom— you want to protect yourself from, what information you hold that is genuinely valuable, and what would happen if it leaked. The plan comes from there; without it, any measure is smoke.
02
Exposure audit: what is known about you right now
We analyse what information about you already circulates in breaches and open sources: emails, phone numbers, addresses, old passwords, data uploaded by third parties without you knowing. It’s the part that impresses clients most, because there’s usually far more than they expected. From there the work starts: remove what can be removed, and neutralise what can’t.
03
The underlying problem: everything is linked
A single email with your full name on a hundred services. A single phone number for the bank, for Amazon and for whatever app of the day. A single browser for work, banking and everything else. When one of those services suffers a breach —and they all do eventually—, whoever collects that data can pull the thread and reconstruct you entirely.
Our job is to cut those threads: email aliases, numbers segmented by use, unique passwords that don’t identify you, browsers and profiles separated by activity, device encryption. Compartmentalise. So a leak is a leak — not a map of your life.
04
Protection against fraud and social engineering
The more public information exists about someone, the easier they are to deceive. Today’s scams are no longer emails full of typos: they’re calls that appear to come from your bank’s real number, made by someone who knows your name, your recent transaction and your account manager’s name. With AI, voices are cloned from very little material. We work both angles: reducing the information available, and training the person (and their family or their team) to recognise the pattern.
05
Communications: content is not the only thing that matters
A message can be encrypted and still reveal a lot: who talks to whom, at what time, from where, how often. That’s metadata, and in most apps it isn’t protected. We set up channels suited to each kind of conversation —the trivial where you’re comfortable, the sensitive where it’s genuinely protected— and we explain what each one protects and what it doesn’t.
06
Devices and systems
App permission review (that torch doesn’t need your microphone), profile isolation, encryption, a VPN chosen with criteria —open source, audited, no logs, own servers— and, for those who need it, hardened mobile operating systems. All with open, auditable software: if you can’t see the code, you’re trusting blindly.
07
Privacy and AI use
More and more people pour their work, their finances and their entire life into an AI assistant. It’s the biggest personal data leak of recent years and almost nobody perceives it as such. We define what can be given to an external AI and what can’t, with separation between personal and professional identity. And when volume or sensitivity justify it, the alternative is sending nothing out at all: that’s where our AI Engineering service comes in, with models running on your own infrastructure.
08
For companies
The same applied to an organisation: segmentation of identities and access, data policy, team training and reduction of the exposed surface. Most incidents don’t start with a technical failure, but with a person somebody knew a little too well.
What you get
A report of your current exposure, a plan adjusted to your real situation, the configuration implemented and working, and the criteria to maintain it — because this is not an installation, it’s a way of operating.
Let’s talk about your situation
Tell us your case — what worries you and what’s at stake — and we’ll tell you where to start.