Product · Aegis
Aegis, the manager who is accountable.
Aegis coordinates the AI specialists, talks to you from any screen and decides what goes up to the board. It shows each step as it works and records who approved what.
- Every company isolated from day one
- Credentials can stay on your network
- Every action logged, with who approved it
- Trust page
The team
Eight specialists, one manager.
Each specialist has a job title, not a character. On the “Your team” board, you see what each one does, its autonomy level and what it did in the period.
Inventory
Knows what the company has, who owns it and what is exposed. Flags unknown machines and handles the outside view.
Vulnerabilities
Reads vulnerabilities and pentest reports, points out what to fix first and writes the step by step for the fix.
Monitoring
Triages alerts from every source, all the time, and only calls when it matters.
Incident
Opens the record, investigates, measures the impact and drafts the communication. A person accepts item by item.
Backup
Checks every morning that each critical machine has a verified copy and runs the restore test.
People
Trains and tests employees against email scams.
Third parties
Answers your customers’ security questionnaires, with the evidence for each answer.
Compliance
Reads evidence, suggests where to attach it and calibrates controls based on it.
Autonomy
You choose how much it does on its own.
The level is chosen by the admin, per specialist, and every change goes into the trail.
- Always
Notifies
Notices go out at any level: a certificate about to expire, a copy that stopped.
- Level 1
Suggests
Says what to do and how. Nothing is prepared or done. This is the default.
- Level 2
Prepares and asks for approval
Gets the action ready, and a person approves with one click.
- Level 3
Does it and notifies
Executes on its own only what can be undone, notifies and offers “Undo”.
- What cannot be undone never runs on its own: it stops at level 2 and waits for a person
- Everything you ask for in the conversation waits for your confirmation, at any level
- All AI work goes into the immutable trail, with the specialist as author and who approved it
The “Today” queue
What to do first, in risk order.
The daily queue opens with Aegis’s morning summary and splits everything into now, this week and later. Each line says what, why it matters and the next step.
- What has a legal deadline
- The incident in progress
- The gap an attacker would use today
- The backup that does not protect
- The dangerous access
- Deadlines
- Everything else
When IT marks something as done, the platform confirms it at the next collection.
Monitoring and incidents
Someone is watching, and only calls when it matters.
Alert hub
Alerts from every source in one place. AI groups what belongs to the same attack, says whether it is dangerous, suspicious or noise, and explains in plain language.
Serious alert
Goes by email to the on-call contacts you define, with what happened, where, why it is serious and up to three steps. It never carries passwords or personal data.
Incident record
An immutable timeline, calculated priority and the legal deadlines running. Closing it requires a root cause and proof.
Reporting back
AI writes, it does not make up numbers.
The numbers in the reports come from a calculation without AI. A stopped source becomes “I don’t know”, never a guess.
Morning summary
At the top of the “Today” queue, what the team did overnight and what is waiting for you.
Weekly summary
Every Monday, in five lines: what happened, what the team solved, what is waiting for a person, the riskiest item and what changed in the risk.
Monthly report
On the first business day, by email and as a PDF: risk in money terms and how much it changed, what each specialist did and up to three decisions for the board.
Limits, in writing
What Aegis does not do.
- It does not invent actions without findings, nor numbers without data
- It does not execute on its own what cannot be undone
- It does not notify the data protection authority for you: it shows the deadline and drafts the communication
- It does not replace your IT for physical work: it prepares, your team executes
- It does not send account and machine names to the AI during triage, and the customer’s documents are not kept by the provider
Frequently asked questions
About Aegis and the AI team.
Does Aegis make decisions on its own?
Only at the autonomy level you choose, and only what can be undone. What cannot be undone, such as answering a questionnaire, always waits for a person.
What language does it answer in?
The reader’s language: Portuguese, English or Spanish.
How do I know what the AI team did?
Through the trail, where each piece of work shows the specialist, who approved it and when, through the “Your team” board and through the weekly and monthly summaries.
What from my company goes to the AI?
The minimum for the work. Triage sees “privileged account”, not the account name; questionnaires carry no person or machine names; and the documents you upload are processed without being kept by the provider.
Put Aegis to work in your company.
Book the assessment: the team starts with the snapshot and the first actions in risk order.