Portugal · live

DOGER: Sovereign decision intelligence for European data.

Agentic decision-support intelligence system: for citizens, businesses, and government.

Property Distressed assets Wildfire Statistics Public contracts Economy Livability

Request demo access →

How it grows

A self-expanding agentic decision-support system.

Public data from every agency, queried as one. Add a source or a use case, and the agents grow with it.

Data we work with

Sovereign by design

EU-hosted, audit-logged, deployable inside a buyer's environment. No personal data unless the buyer scopes it in.

Cross-domain by default

One agent answers across property, wildfire, statistics and the economy in a single reply, with every figure cited to its source.

What it can do

Answer multi-source questions.

Deterministic tools per domain Every figure cited Checked calculations

AGT Agent

Answers one natural-language question across all of the below.

RE Property

Portal and bank listings, auctions · deduplicated

ZON Zoning

Zoning at any point · sites ranked for a use

FIRE Wildfire

Fire history · 9-day probability forecast

WX Weather

Active warnings by district · local conditions

STAT Statistics

Population · housing prices · crime · tourism

ECON Economy

Incomes · jobs · businesses · insolvencies

PUB Public money

Council revenue · who won which contracts

LIV Livability

Air quality · commute · transit · amenities

DOGER Intelligence answering how population and the 65+ share changed in Braga, Coimbra and Faro, 2015 to 2025 — Official statistics · one evidenced answer · all 308 municipalities

"How likely is a wildfire on my Alentejo parcels over the next 9 days — and which municipalities are highest-risk this week?"

Wildfire forecast · 9 days ahead · mainland Portugal · refreshed daily

Request demo access

For governments · investors · infrastructure developers

"Where can the next industrial site go?"

Every mainland municipality ranked on siting fit, in one query.

DATA Data joined

Six national datasets joined per municipality. Nobody publishes them together.

6 datasets · 278 mainland municipalities

ANS What it answers

A site-selection brief comes in. The agent can join six datasets per municipality, rank every mainland municipality on land zoned for industry, planning constraints, wildfire exposure, land prices, labour pool and municipal finances, and return a top-ten shortlist.

Example ranking shown · the answer cites every factor per site
Re-weight the criteria and re-run the ranking

For insurers · landowners · civil protection

"Which estates carry summer wildfire exposure?"

Wildfire probability on each parcel for the next 9 days.

DATA Data joined

Four datasets joined per parcel: fire history, your parcel locations, the daily fire forecast and live weather warnings.

4 datasets · forecast refreshed daily

ANS What it answers

A landowner sends seventeen parcels across Alentejo: 8 cork stands, 6 olive groves, 3 vineyards. The agent returns each parcel's fire probability for every day from d+1 to d+9. Most cork stands come back red, most olive groves green.

Evidence for pricing parametric fire cover on the asset

For banks · regulated lenders

"What is this loan actually secured on?"

Collateral value and wildfire exposure, with the full audit trail.

DATA Data joined

Four datasets joined on each pledged property: market comparables, distressed listings, wildfire exposure and zoning.

4 datasets · keyed on the property

ANS What it answers

A bank underwrites a €2M SME loan in Castelo Branco. The agent locates the four warehouses pledged as collateral, values them against asking prices and bank-owned listings nearby, checks their zoning, and flags that three sit in the highest wildfire-probability band.

Evidence on the asset for the credit file · full audit trail
Every line cited to its source · the credit decision stays with the bank

For family offices · private banks

"What is my client's Portuguese portfolio really worth?"

Consolidated portfolio · valuation · wildfire overlay · one due-diligence pass.

DATA Data joined

The client's own list of holdings joined to three public datasets.

Client list + public records

ANS What it answers

A private banker, with client consent, loads a 14-property portfolio. The agent values each asset against comparable listings, sets it against local price and tourism trends, checks its zoning, and surfaces wildfire exposure on three parcels.

One due-diligence pass · single investment report · audit trail
The client's own list, used with consent

Closed beta · Portugal first

Your data Our solution within a month

Closed beta. Sovereign infrastructure. Describe your problem — we'll ship the solution.

Request demo access →

Built by the DogerAI team — public-procurement anomaly detection on the full Prozorro corpus since 2022.