Academy

Welcome to Analytica's Academy.
Here you will find all the documentation and information to deeply understand how our platform works. You will also find a set of tutorials to understand how to create your Agentic AI.

Agentic AI

Analytica is based on a set of Agents, which are designed to perform specific tasks, more or less complex, coordinated by an LLM.
Through our system, it is possible to translate commands, more or less simple, into the execution of different agents to achieve a certain result. We are no longer talking about a simple GPT, but a much more complex system.

Our Agent system is at the core of every functionality: it responds to chat, but also executes workflows, and prepares responses to your tickets.

The Knowledge Base

The Knowledge Base is a fundamental part of Analytica. It is the place where your information lives, already structured and ready to be used.
It should not be confused with a set of documents. Those are files. Our Knowledge Base already contains the information EXTRACTED from files (and from other data sources). It is much more similar to an actual database, in the IT sense.

More information is available on the Knowledge Base page.

Chat

The Chat is the section that allows you to interact with Analytica and its agents in a conversational way.
Based on the request or message entered, the various agents will be executed to perform the task.
Within the chat, it is possible to view data in tabular format or with charts. The formatting of this data will always be customizable at will, with a simple text command.

Depending on the type of installation, it is possible to transcribe your commands using the microphone as input.

More information on how the chat works is described on the dedicated page.

File Upload in Chat

The chat also supports file upload. Unlike classic AI tools, such as Gemini or ChatGPT, files uploaded in chat can be managed in two ways:

  • Chat Only, for classic behavior: the file will be processed, and its content will be available for searches or queries, exclusively within the chat used;
  • Knowledge Base, and therefore the file content will be cataloged, extracted, and inserted into the knowledge base.

By selecting Knowledge Base, the content will therefore potentially be made available to other users as well, and will be structured within the Knowledge Base.
If the purpose is only to query and process a file (for example, I want a summary of a certain contract), then the correct method is Chat Only.

Dashboard

The Dashboard allows the creation of views on your data.
Each dashboard is made up of widgets.
A widget is a visualization of certain data.
There are two types of widgets:

  • Knowledge Base, which allows you to execute a query on the Knowledge Base, displaying the data in a table;
  • Prompt, which instead executes a command (i.e. an actual prompt) and displays its result.

It is possible to modify the number of widgets, their layout and their size.

Through the prompt, it is possible to create charts or tables, with data coming from the execution of agents.

Files

The Files section allows you to review and monitor the current files present on the platform.
In the file list, in addition to the original name of the uploaded file, its Knowledge Base will be described.
The Knowledge Base indicates how the file has been classified according to the various types defined in the Knowledge Base section. By pressing the "info" icon, it is also possible to inspect the various properties that have been identified and structured from the original file.

The files available in this section come from:

  • direct upload and upload
  • insertion of files within the chat
  • synchronized folders

Synchronized Folders

In this section, it is possible to define remote folders, through connectors for Google Drive and NextCloud.
The content of these folders can therefore be imported directly into Analytica, and will then be displayed in the Files section.
Each imported file will then be classified according to the types defined in the Knowledge Base.

Workflow

Workflows are commands and prompts that are executed following a specific trigger.
The trigger can be temporal (for example, every day at 08:00), manual (I manually start the workflow), or from a change of data within the Knowledge Base (a new Transfer is inserted).

Through workflows I can therefore perform tasks automatically, making Analytica reactive.

A typical example of Workflow can be the generation of a report and its sending via email, or the execution of data import from a website.