---
title: Models
description: How Avra's two foundation models compare — Avra RFM for your own relational data, Avra GFM for Brazilian companies — and which fits.
---

Avra has two foundation models. [Avra RFM](/rfm) reads your own relational database and predicts the outcomes you define. [Avra GFM](/gfm) reads Avra's knowledge graph of the Brazilian economy and returns signals for any Brazilian company. Both reach your systems through [Avra Core](/core), live or in batch, and Avra RFM can also run inside your AWS account.

<svg class="fig" viewBox="0 0 640 260" role="img" aria-label="Two rows. In the first, your tables go into Avra RFM, which returns scores and estimates for the outcomes you define. In the second, a company's CNPJ goes into Avra GFM, which reads Avra's knowledge graph and returns probabilities of default and embeddings.">
  <defs>
    <marker id="fig-models-arrow" viewBox="0 0 10 10" refX="9" refY="5" markerWidth="7" markerHeight="7" orient="auto-start-reverse">
      <path d="M0,0 L10,5 L0,10 z" class="arrow" />
    </marker>
  </defs>
  <text class="small title label" x="16" y="24">You bring</text>
  <text class="small title label mid" x="307" y="24">The model</text>
  <text class="small title label" x="424" y="24">You get back</text>
  <rect class="card" x="16" y="46" width="172" height="48" rx="8" />
  <text class="small title" x="32" y="66">Your tables</text>
  <text class="small label" x="32" y="83">as they are, any schema</text>
  <line class="edge" x1="188" y1="70" x2="226" y2="70" marker-end="url(#fig-models-arrow)" />
  <rect class="solid" x="232" y="42" width="150" height="56" rx="10" />
  <text class="mid on-solid title" x="307" y="75">Avra RFM</text>
  <line class="edge" x1="382" y1="70" x2="418" y2="70" marker-end="url(#fig-models-arrow)" />
  <rect class="soft" x="424" y="46" width="200" height="48" rx="8" />
  <text class="small title" x="440" y="66">Scores and estimates</text>
  <text class="small label" x="440" y="83">for the outcomes you define</text>
  <rect class="card" x="16" y="146" width="172" height="48" rx="8" />
  <text class="small title" x="32" y="166">A company's CNPJ</text>
  <text class="small label" x="32" y="183">any Brazilian company</text>
  <line class="edge" x1="188" y1="170" x2="226" y2="170" marker-end="url(#fig-models-arrow)" />
  <rect class="solid" x="232" y="142" width="150" height="56" rx="10" />
  <text class="mid on-solid title" x="307" y="175">Avra GFM</text>
  <line class="edge" x1="382" y1="170" x2="418" y2="170" marker-end="url(#fig-models-arrow)" />
  <rect class="soft" x="424" y="146" width="200" height="48" rx="8" />
  <text class="small title" x="440" y="166">Probabilities of default</text>
  <text class="small label" x="440" y="183">and embeddings</text>
  <line class="edge" x1="307" y1="220" x2="307" y2="202" marker-end="url(#fig-models-arrow)" />
  <rect class="card dash" x="212" y="222" width="190" height="28" rx="14" />
  <text class="small mid label" x="307" y="240">Avra's knowledge graph</text>
</svg>

Avra RFM reads your own tables and predicts the outcomes you define. Avra GFM starts from a company's CNPJ and reads Avra's knowledge graph around it.

## How they compare

| | Avra RFM | Avra GFM |
|---|---|---|
| **What it is** | The foundation model for relational data | A graph-based foundation model, specialized for Brazil |
| **Reads** | Your relational database, any schema | Avra's knowledge graph of the Brazilian economy |
| **Starts from** | Your tables, keys, and history | A company's CNPJ |
| **You receive** | A score for each record, for a yes-or-no outcome, or an estimate, for a number | Probabilities of default at five horizons, from 30 days to a year, and a 1024-dimensional embedding |
| **Specialized for you** | Reads your history in context by default, and a copy can be adapted or fine-tuned for your workspace | A credit model built for your portfolio on its embeddings, your features, and your outcomes |
| **In Avra's cloud** | Through Avra Core, live or in batch | Through Avra Core, live or in batch |
| **In your AWS account** | On Amazon SageMaker, through AWS Marketplace | Runs in Avra's cloud only |

## Which model fits

Start from where the answer lives: in your own records, or in the context around a company.

| You want to | Use |
|---|---|
| Predict an outcome you define, such as churn, delinquency, conversion, or fraud, for the records in your own tables | [Avra RFM](/rfm) |
| Estimate a number from your own history, such as how much each customer will spend next month | [Avra RFM](/rfm) |
| Know a Brazilian company's probability of default, at horizons from 30 days to a year | [GFM probabilities of default](/use-cases/credit-intelligence) |
| Give the models you already run the relationships around each company | [GFM embeddings](/gfm/embeddings) |
| Learn your own definition of default from the GFM's view of each company | [A credit model on GFM embeddings](/gfm#make-it-yours) |

The two work side by side. A lender can predict delinquency over its own loan book with Avra RFM, and read Avra GFM's probabilities of default for the companies applying, both through the same API.

## How you reach them

Both models reach your systems through [Avra Core](/core): one record at a time through the [live API](/core/predictions) when a decision comes up, or a whole list in [batch](/core/batch-inference) on your schedule. Each model has versions, and an alias such as `production` decides which one answers, so your systems move to a new version only when the alias does. See [Model versioning](/core/model-versioning).

Avra RFM can also run inside your AWS account, on Amazon SageMaker. See [Avra RFM on AWS SageMaker](/deployment/aws-sagemaker).

## Where to go next

**[Avra RFM](/rfm)**

Predict any outcome from your own relational data.

**[Avra GFM](/gfm)**

Read every Brazilian company through the relationships around it.

**[Avra Core](/core)**

Call either model live or in batch, with versions and webhooks.

**[Use cases](/use-cases)**

Both models at work in credit, fraud, and acquisition.
