<aside> <img src="/icons/info-alternate_green.svg" alt="/icons/info-alternate_green.svg" width="40px" /> This overview will introduce you to the platform’s key concepts and how they work together to create powerful automation workflows.
</aside>
Neuron Builder is where you create your AI agents by assembling building blocks called Axons. Axons represent individual tasks or steps the agent performs. The Neuron Builder is designed to make this process intuitive—simply drag and drop the necessary Axons from the Axon Library and connect them to build your Agent. Each Axon plays a specific role, whether it’s gathering data, processing it, transforming it, or sending it to the desired location.
The Axon Library contains a collection of Axons divided into four main types: Inputs, Processors, Transformers, and Outputs. Each Axon type serves a unique purpose in the agent’s workflow, and there are specific rules for how they connect to each other.
Input Axons are responsible for bringing data into the agent. They represent the sources from which the agent retrieves or monitors data. Common input Axons include:
• API Connector: Integrates with external APIs to fetch data.
• Gmail Reader: Fetches and filters emails based on predefined criteria.
• SharePoint Input Connector: Monitors SharePoint for file uploads or changes.
• Table Uploader: Uploads structured data like spreadsheets or CSV files.
• Uploader: Allows users to manually upload files to be processed.
Inputs can only connect to Processors, as they supply the data the agent will act on.
<aside> <img src="/icons/info-alternate_green.svg" alt="/icons/info-alternate_green.svg" width="40px" /> Inputs can only connect to Processors.
</aside>
Processor Axons handle the data processing that takes place once the agent receives input. These Axons include various tools for analyzing, classifying, extracting, and manipulating data. Some key Processor Axons are:
• Classifier: Automatically classifies documents or data based on labels provided.
• Comparator: Compares incoming documents or data against templates to highlight differences.
• Decision Tree: Adds conditional logic, allowing the agent to take different actions based on specific criteria.
• Extractor: Identifies and extracts specific information from documents or datasets.
• Content Generator: can be used to generate custom content using information and results from previous axons.
<aside> <img src="/icons/info-alternate_green.svg" alt="/icons/info-alternate_green.svg" width="40px" /> Processors can connect to each other, to Transformers, or directly to the Database Axon.
</aside>
Transformer Axons are responsible for modifying or reformatting data before it is sent to its final destination. They take the processed data and apply final touches, such as merging, renaming, or generating files. Key Transformer Axons include:
• Doc Generator: Fills document templates (e.g., Word or PDF) with extracted data.
• File Merger: Combines multiple files into a single document.
• File Renamer: Renames files based on predefined rules.
<aside> <img src="/icons/info-alternate_green.svg" alt="/icons/info-alternate_green.svg" width="40px" /> Transformers can connect to other Transformers or to Outputs.
</aside>
Output Axons are the final step in the workflow. These Axons define where and how the processed data is sent or stored. Some common Output Axons include:
• Google Drive Output Connector: Saves processed files to Google Drive.
• SharePoint Output Connector: Transfers processed files to SharePoint Document Center.
• Database Connector: Stores extracted data in a database for further analysis or reporting.