Skip to content

AI and privacy, built in Brazil

AI for business.
Privacy before inference.

Sanitize sensitive data, connect models and track usage. Your organization's policy, built into the flow of data.

30 days free. Create your account and try it in the browser. No card, with a starter quota.

  1. Your data
  2. Privacy layer
  3. AI and integrations

Privacy you can see

See what changes before it moves on.

Fictional examples · precomputed results
Contact and support

Contact and support

01 / 10

Original text

Fale com Carla Oliveira pelo e-mail carla.oliveira@example.invalid ou pelo telefone (21) 98888-7777.

Preview result

Fale com [PESSOA_1] pelo e-mail [EMAIL_1] ou pelo telefone [TELEFONE_1].

Highlighted passages were replaced with placeholders.

Registration with a CPF

Registration with a CPF

02 / 10

Original text

Paciente Bruno Pereira, CPF 529.982.247-25, em acompanhamento.

Preview result

Paciente [PESSOA_1], CPF [DOCUMENTO_1], em acompanhamento.

Highlighted passages were replaced with placeholders.

Healthcare

Healthcare

03 / 10

Original text

A paciente Ana Souza aguarda o resultado do exame.

Preview result

A paciente [PESSOA_1] aguarda o resultado do exame.

Highlighted passages were replaced with placeholders.

Delivery and address

Delivery and address

04 / 10

Original text

Entregar a Renata Conceição na Rua das Acácias, 120, São Paulo.

Preview result

Entregar a [PESSOA_1] na [RUA_1], [NUMERO_1], São Paulo.

Highlighted passages were replaced with placeholders.

Two people

Two people

05 / 10

Original text

Maria Clara de Oliveira encaminhou o pedido para José Antônio Pereira.

Preview result

[PESSOA_1] encaminhou o pedido para [PESSOA_2].

Highlighted passages were replaced with placeholders.

Message with an emoji

Message with an emoji

06 / 10

Original text

📋 Paciente: Márcia D'Ávila. Evolução clínica favorável.

Preview result

📋 Paciente: [PESSOA_1]. Evolução clínica favorável.

Highlighted passages were replaced with placeholders.

Repeated name

Repeated name

07 / 10

Original text

Rafael Nascimento abriu o chamado. Rafael Nascimento aguarda retorno.

Preview result

[PESSOA_1] abriu o chamado. [PESSOA_2] aguarda retorno.

Highlighted passages were replaced with placeholders.

Structured data

Structured data

08 / 10

Original text

{"nome":"Roberto Almeida","observacao":"sem sintomas"}

Preview result

{"nome":"[PESSOA_1]","observacao":"sem sintomas"}

Highlighted passages were replaced with placeholders.

Clinical context

Clinical context

09 / 10

Original text

Diabetes mellitus tipo 2. Hemoglobina glicada 7,2%. Manter metformina 500 mg e reavaliar em três meses.

Preview result

Diabetes mellitus tipo 2. Hemoglobina glicada 7,2%. Manter metformina 500 mg e reavaliar em três meses.

No personal data detected in this example. The text was kept unchanged.

Public information

Public information

10 / 10

Original text

São Paulo e Belo Horizonte registram crescimento na demanda por tecnologia.

Preview result

São Paulo e Belo Horizonte registram crescimento na demanda por tecnologia.

No personal data detected in this example. The text was kept unchanged.

Control at every step

Your policy. Across your integrations.

Sanitize before sharing

Replace data with placeholders, redact or block content according to your organization's policy.

CPF / CNPJEmailPhone
Explore the rules

Auditing with restricted access

Track usage and detected categories. Content retained for auditing is encrypted, with a defined retention period.

AES-256-GCMOrganization policies
Understand auditing

Evidence you can evaluate

Explore methods, tests and limitations. Use examples from your workflow to evaluate successes and failures before integrating.

BenchmarksDocumented rules
Explore the results

Platform usage

Aggregate API usage

Requests
—processed in production
Personal data
—recorded occurrences

Built to integrate

Your systems. One privacy layer.

Use the API to sanitize data before sending it to your chosen destination. Or connect your application to Rota models through an OpenAI-compatible API.

Text and JSON
Synchronous sanitization and batch processing.
Audio and documents
Transcription and text extraction. Image OCR in beta.
Usage and governance
Keys, quotas, policies and records in your dashboard.
Read the docs
POST /v1/privacy/cleancurl
curl https://rota-nacional.ia.br/v1/privacy/clean \
  -H "Authorization: Bearer $ROTA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Contato: teste@example.com",
    "policy": "placeholder"
  }'
Sanitized text · exampleContato: [EMAIL_1]

What's next. A browser extension, desktop app, ready-made connectors and dedicated deployment are planned. See capabilities and roadmap

AI models

One API. A model for every task.

Explore the catalog

Rota Summary

64k tokens

A cost-effective model for summaries, classification, triage and focused tasks at scale.

summaryfastlow costPT-BR

Rota Analysis

128k tokens

A capable model for careful analysis, technical reviews and decisions with broad context.

analysisreasoningPT-BR

Rota Code

131k tokens

A model curated for technical assistance, automation, system reviews and large codebases.

chatcodereasoningPT-BR

Rota Agents

131k tokens

A model for assisted automation, tool use, planning and guided execution.

codeautomationPT-BR

The platform's next chapter

From text to image. From voice to document.

rn.* · in development

A family of capabilities to create, transform and understand content through the same API. Explore the product direction and API contracts in progress.

In development

Conversations and semantic search

Assistants and vectors to find information by meaning.

rn.chatrn.embeddings
View endpoints and example
POST /v1/chat/completions
POST /v1/embeddings

{
  "model": "rota/embedding",
  "input": ["Política de privacidade"]
}

Example request body for embeddings. Chat on RN's fleet is planned with the model rota/local.

In development

Transcription and speech

Turn speech into text and text into speech for your applications.

rn.audio.transcriptionsrn.audio.speech
View endpoints and example
POST /v1/audio/transcriptions
POST /v1/audio/speech

{
  "text": "Bem-vindo à Rota Nacional."
}

Example request body for speech. The transcription API already exists; execution through the RN namespace is in development.

In development

Music and composition

Create soundtracks from a description and prepare audio compositions.

rn.audio.musicrn.audio.compositions
View endpoints and example
POST /v1/audio/music
POST /v1/audio/compositions

{
  "style": "Instrumental ambiente",
  "seconds": 60
}

Example request body for music. The composition structure is still being defined.

In development

OCR and document extraction

Convert images and documents into text for analysis and integration.

rn.ocrrn.documents.ocr
View endpoints and example
POST /v1/ocr
POST /v1/documents/fidelity/jobs

Content-Type: multipart/form-data
file: documento.pdf

File upload. Current extraction remains available at /v1/privacy/extract, with image OCR in beta.

In development

Document review

Prepare document review workflows with processing status tracking.

rn.documents.review
View endpoints and example
POST /v1/documents/review/jobs
GET /v1/jobs/{id}
GET /v1/jobs/{id}/result

Planned workflow: submit, track and retrieve the result. The input format is still being defined.

In development

Image generation and editing

Create images from a description or transform an existing image.

rn.images.generationsrn.images.edits
View endpoints and example
POST /v1/images/generations
POST /v1/images/edits

{
  "prompt": "Uma cidade-jardim ao amanhecer"
}

Example for generation. Editing is planned with a file in image and an instruction in prompt.

Integration preview. These examples document work in development, subject to change and validation. They do not indicate availability in the free trial or release for production. Requests will require authentication and access to the capability.

About the company

Developed by Exponencial.

Exponencial, the company behind Rota Nacional, is a NVIDIA Inception member and participates in AWS Activate.

Explore the investment opportunity
NVIDIA Inception — startup program NVIDIA Inception Global program for technology startups AWS Activate Startup support program Access through NVIDIA Inception

Start with a text. Build toward your workflow.

Your next project starts here.

Try it in your browser or talk to the team about integration, volume and policies.

Try it free

FAQ

Frequently asked questions.

Are rn.* capabilities already available?

The rn.* namespace groups capabilities in development for chat, embeddings, audio, OCR, documents and images. Endpoint examples are an integration preview and may change. Current transcription and extraction remain available through their documented routes; this does not mean the new RN execution is released. Contact the team before planning production use.

What does Rota do with the data?

It identifies personal data in submitted content and applies your organization's policy: placeholders, removal or blocking. The dashboard records categories and usage. Detection has limitations; results need to be evaluated for the process in which they will be used.

Do I need to integrate a system to try it?

No. You can create your account and test a text in the browser. To apply the processing to an automated organizational process, your technical team needs to integrate the API. Ready-made extensions and connectors are still planned.

What data does the current version detect?

Person names are identified by our own neural engine, running in São Paulo. Email, phone, Brazilian identifiers and other categories go through rules; numeric CPF and CNPJ pass check-digit validation. See the limits in the engineering documentation.

Where can I inspect benchmarks and papers?

On the engineering page, we publish results from 46 synthetic texts, remaining errors, annotated data, predictions and a scoring script. The same page maps papers and standards to the techniques used. This is experimental evidence, without certification or guaranteed production coverage.

Are the examples on this page processed live?

No. These are ten fictional Portuguese texts with precomputed results from the same name engine used by the API. You can compare the examples without sending data. The results do not guarantee the same coverage on other texts.

What does the trial cost, and what happens afterwards?

The trial lasts 30 days, with a starter quota and no card. After that, usage continues with prepaid credit in reais via Pix, from R$ 5. For volume, integration or a contract, talk to the team.

How do I sign in to try it?

Registration is open: sign in with email, Google or a passkey and accept the terms. The dashboard test uses the same processing as the API: the name engine and the rules.

Is submitted content stored?

Registration explains the terms and encrypted retention of input and output. When auditing is enabled, this content may be retained for the specified period with restricted access. Standard usage records show metadata, such as categories and counts, without raw text.

Does data stay exclusively in Brazil?

Operations and curation are Brazilian, but the current service does not guarantee that all data and inference remain exclusively in Brazil. Dedicated deployment, jurisdiction-specific residency and offline operation require a specific project and validation.

Does sanitization eliminate every exposure risk?

Detection reduces exposure but cannot eliminate every risk. Monitoring policies do not sanitize data; OCR is in beta. Compliance also depends on the organization's configuration and processes.

Do I have to use Rota's AI models?

No. Your team can use the API solely to sanitize data and send the result to its chosen destination. Rota also provides access to AI models if that suits your integration.