# Solution blueprint

## **Introduction**

This is a high-level diagram of our solution, detailing the primary modules and the steps involved in processing and accessing the information.

## **Platform Architecture Overview**

<figure><img src="/files/3zpclU0wmXKZvuwvYjVe" alt=""><figcaption></figcaption></figure>

**AI Data Processor Module:** This module is responsible for interpreting the uploaded files and extracting the relevant information. The type of information extracted depends on the file type; currently, we support utility invoices exclusively. Our development efforts are focused on expanding capabilities to include certificates, purchase orders, invoices, and more complex documents.

**Data Validation & Heuristic Module:** This module executes various heuristics and validations on the information extracted from documents. Its primary objective is to ascertain the accuracy of the information and maintain consistency across all processed documents.

**Auditable Storage Module:** This module manages the storage of all processed information. It creates a data point that encompasses the structured information extracted and the original document from which this information was sourced. At the moment, we handle utility data points that encompass electricity, gas, and water utilities. Any alterations to a data point are meticulously tracked to document what changes were made to the information and who made them. Optionally, we can leverage blockchain technology to enable public auditing of this information.

**Audit View:** Data Points are accessible via our **Document AI API** or through a link that generates an HTML page. This feature is particularly beneficial for users who wish to make the information available for auditing purposes. The Audit View displays the entire change history of a data point, along with the original document.

## **Infrastructure**

Our infrastructure is divided in 2 main components:

* Studio / API: This piece of technology is running in **fly.io** and its build with **Phoenix** and **Elixir**.
* Our AI module use **Open AI API**s and we use the **AWS Textract** Service.
* Tesseract OCR and Google Document AI OCR *(coming soon)*

<div><figure><img src="/files/OF7SUUEpJjSFEXjlHLwD" alt="" width="188"><figcaption></figcaption></figure> <figure><img src="/files/xA85iVl1KRB089CLPKXc" alt="" width="375"><figcaption></figcaption></figure> <figure><img src="/files/PK6rV22KK6rcEVcE1wC9" alt="" width="375"><figcaption></figcaption></figure></div>

## Authentication

We design our APIs for backend access by our users. To utilize these, you must have a valid **access token**. You can choose the duration for which the token remains valid; available options are 30, 60, or 90 days. This token allows you to execute all operations and access all our endpoints. You may also revoke the token at any time.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.cedalio.com/tecnology/solution-blueprint.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.