We are building the analysis layer that exists inside financial institutions and does not exist outside them.
Two products in operation, a deterministic calculation engine and revenue that does not depend on transactions. This page explains the thesis. The numbers are in the material.
The analysis layer as software, by subscription.
A good financial decision depends on an analysis layer: cross-referenced and dated data, alternatives on the same scale, opportunity cost calculated, explicit priority. Institutions have had that layer for decades. Outside them, the individual investor, the developer, the chain, the indebted family; it simply does not exist.
It is not a problem of access to information. Information is abundant. It is a problem of translation: nobody turns the available data into a defensible decision at the moment and in the language of whoever decides.
We build that layer as software, by subscription, for both sides of the balance sheet. Capital allocation on one side, liability management on the other. The same calculation discipline serves both, and serves the next markets we take on.
Why now.
Brazilian public data became usable.
Property registry, economic registry, safety and mobility data are available in a volume and granularity that did not exist five years ago. They remain fragmented, which is exactly the opportunity.
Language models solved translation, not calculation.
Explaining a result in natural language became cheap. Producing the right result remains expensive and hard. Whoever treats an LLM as a calculation engine delivers numbers that do not hold; whoever treats it as an explanation layer over a deterministic engine delivers an auditable product.
Subscription distribution reached the right audience.
The Brazilian retail investor already pays for analysis tools in other markets. They do not pay for real-estate analysis because it does not exist in their format.
The two markets.
Capital allocation
Investment decisions that depend on a place: which region, for which use, over which horizon, at what maximum price.
The engine assesses a territory across eight dimensions, return, trend, demand, search pressure, economic fabric, infrastructure, safety and legal risk; with a weight matrix and analysis radius of its own for each of seven intended uses. The same region receives different scores depending on what is intended there, which is the opposite of what a market index delivers.
It delivers a year-by-year proforma from 1 to 15 years with full costs, IRR, NPV against the benchmark rate, payback, optimal exit year and break-even. Maximum offer by reverse valuation. Comparison between territories on the same scale.
Liability management
The way out of debt with a defined route and weekly tracking.
The engine calculates free cash, classifies urgency into three bands and elects among three amortisation strategies by deterministic rule with explicit disqualifications. The queue ordering respects a survival hierarchy before any interest optimisation: risk of civil imprisonment, housing, repossession, judicial collection and essential service.
When a debt is paid off, the amount cascades automatically to the next in the queue.
The next markets follow the same selection criterion: relevant capital at stake, decisions taken without analysis, and data available enough to produce a defensible answer.
Defensibility.
The rule is the asset, not the data.
Anyone can access a public source. The weight matrix per intention, the priority hierarchy, the eliminating factors and the confidence rules are accumulated calibration; and every round with real users improves it.
Determinism creates auditability, and auditability creates trust.
A competitor who generates a number from a language model cannot explain how it got there. We can, line by line, and that is what makes it possible to sell to those who move relevant capital.
The absence of transactional revenue is a position, not a limitation.
A listings portal and a credit originator cannot deliver neutral analysis without cannibalising their own revenue. We can, and it is the one thing they cannot copy without restructuring.
Local depth before geographic breadth.
Calibrating a region requires field work that does not scale by scraping. That is slow for us and equally slow for whoever comes next.
The product in operation.
Long-term rental
78/100
Data confidence: high- Demand+18
- Economic fabric+11
- Infrastructure−6
Residential subdivision
41/100
Data confidence: medium- Legal risk−21
- Infrastructure−9
- Trend+7
Seven independent weight matrices, calibrated by intention. The score is not a weighted average of a single index.
Hover the chart to see the values for each year.
Full costs built in, transfer tax, deed, registration, brokerage, maintenance, vacancy and income tax. Comparison against the benchmark rate is mandatory in every projection.
| Debt | Balance | Monthly interest | Reason for the position |
|---|---|---|---|
| Rent in arrears | 4.200 | — | housing |
| Car loan | 18.400 | 1,9% | repossession |
| Credit card | 6.900 | 12,4% | highest interest |
| Personal loan | 11.300 | 3,1% | smallest remaining balance |
The ordering respects a table of risk weights before any financial optimisation. It is a product rule, not a heuristic.
Revenue model.
Recurring subscription, three tiers per product, with a free entry plan and conversion by usage. No revenue comes from commission, advertising, credit origination or data sales.
In the capital allocation product, alongside the individual subscription there is a multi-user plan and bespoke studies for land developers, property developers and chains. Leads with a signal of urgency go straight to scheduling.
Current prices, traction, cohort and conversion metrics are in the material sent after contact.
If the thesis makes sense, write.
We reply with the full material, deck, traction numbers and terms; after a quick screening. Include in the email:
- Name and organisation
- Type of investor: angel, fund, strategic or other
- One sentence about your interest
Request access to the material
We do not make a public offering of securities. Contact is individual and the material is sent after assessment.