How exposed are we?
- Today's answer
- A maturity score from 1 to 5, which fits no budget line.
- Indene's answer
- Expected loss in currency, as a range, with confidence beside it.
Control assessment with an evidence hierarchy, quantified into financial exposure that holds up in the boardroom and in front of the regulator.
Expected annual loss
R$ 8.4M to R$ 21.7M
Example figures: they exist to show the output format, not an actual measurement.
The three questions
Why now
Published 18 Dec 2025: 14 verifiable minimum procedures and controls, an independent annual intrusion test, and evidence retained for 5 years.
Resolutions CMN 5,274/2025 and BCB 538/2025
The adaptation deadline has passed. Institutions still answering with spreadsheets are answering late.
CMN 5,274/2025
The Central Bank's supervision staff fell from 1,100 to under 650. Inspection arrives as a written request; whoever responds with a dated, traceable dossier has an objective advantage.
Sinal/BC, 2025
Average cost of a data breach in the Brazilian financial sector. It is the argument the board understands without translation.
IBM/Ponemon, Cost of a Data Breach 2025
How it works
The product path is the onboarding path: a first exposure reading within 10 business days.
Every control gets two scores, design and effectiveness, that never collapse into a single number.
Evidence arrives via documents, CSV and connectors. Each with origin, date and author.
The FAIR engine with Monte Carlo translates control state into a loss range, with confidence pinned to the number.
A queue ordered by exposure reduction, and an inspection dossier as PDF: immutable trail, no manual collection.
Evidence hierarchy
“When did you last actually restore from backup, time it, and have the result match the declared RTO?”
Someone said so. Deliberately colorless; it proves nothing yet.
A document, a tool export, a connector collection.
An active test that answered. Only this raises the ceiling.
The three metrics
If our systems are encrypted tomorrow, how much do we lose and how fast do we recover?
What data of ours is already leaving through AI tools nobody approved?
If a critical vendor goes down or leaks, how much do we lose, and can we prove to the regulator that we knew?
Simulation
Toggle the controls and watch the range shrink: on the same scale, because comparing ranges on different scales is a visual lie.
Single scale from R$ 0 to R$ 24M on both ranges.
Example figures: they show the output format, not an actual measurement.
One control, every framework
Control-to-requirement mapping is many-to-many, with an explicit degree of coverage. A control assessed once answers every framework mapped to it.
Tested immutable backups
Full and partial coverage are shown separately; partial never rounds up to "met".
What you take away
3 to 5 pages in the language of the board. Opinionated, printable, generated with history.
Origin, date and trail for every piece of evidence, with zero manual collection. The promise is commercial because the database guarantees it.
Where the data comes from
Connected today
In the catalog · not yet verified against a real environment
A connector brings evidenced evidence, never validated; only an active test validates.
Aegis · AI on rails
The AI reads documents and proposes classification, origin and excerpts, on rails, under versioned prompts.
No risk snapshot exists without human sign-off.
A content filter runs before every call: sensitive data is stopped at the rails before it ever reaches the model.
The product measures its own no-edit acceptance rate. The stated target is 70%.
Who uses it
10 minutes before the meeting
Range, confidence, variation and sector benchmark. What to bring to the council.
All day, during inspection season
Program, frameworks, SoA, work queue and filtered-view export.
Gets in, fixes, attaches proof, gets out
Consolidated view, vulnerabilities, backup, detection and compliance.
Design partner program
We have no customer to name yet, and we will not invent one. The first five accounts join at half the list price for the first year, onboarding included, a first exposure reading within 10 business days, and roadmap priority.
A 30-minute demonstration: range, confidence, simulation, and the dossier exported as PDF. A person answers, within 24 business hours.
Or write directly: contato@indene.io