Skip to content
Nappai Documentation

Ai Filter

This component helps you quickly find the information you need within your data by using simple, everyday language. Instead of writing complex code, you can ask questions in plain English to filter your data.

Relationship with AI Language Models

This component uses an AI language model (like those from OpenAI) to understand your requests and filter your data accordingly. The accuracy of the filtering depends on the quality of the language model and how clearly you phrase your request.

Inputs

  • Data: This is the data you want to filter. It can be various types of information, like text, numbers, or even messages.
  • Query: This is where you type your question or request in plain English. For example, you could type “Show me all sales from last month” or “Find all customers in California”.
  • Max Chunks: This lets you specify the maximum number of data pieces the AI will process at once. Leave this blank for automatic handling, or use it to control processing time for very large datasets.
  • Language Model: You need to select the AI language model you want to use for filtering. Nappai will provide options. This is a required field.

Outputs

  • Filter: This output contains the filtered data based on your query. This is the result of your search.
  • Tool: This output provides a reusable tool that encapsulates the filtering logic. This is more technical and may not be needed for basic use.

Usage Example

Let’s say you have a dataset of customer orders. To find all orders from customers in New York, you would:

  1. Connect your customer order data to the “Data” input.
  2. In the “Query” input, type “Show me all orders from New York”.
  3. Select your preferred Language Model.
  4. Run the component.
  5. The “Filter” output will show you a list of orders from New York.

Templates

This component can be used in various workflows within Nappai. You’ll see it in different automation templates, but its configuration will always be similar.

Many components can work with the Ai Filter’s output. For example:

  • Summarizer: Use this to summarize the filtered data.
  • Google Sheet Writer: Write the filtered data to a Google Sheet for further analysis.
  • Data Conditional Router: Route the filtered data based on specific conditions.

(Note: A full list of related components is provided in the technical documentation, but this section focuses on the most commonly used ones.)

Tips and Best Practices

  • Be clear and specific in your query: The more precise your question, the better the results.
  • Experiment with different language models: Different models may perform better depending on your data and query.
  • Start with smaller datasets: If you have a very large dataset, start by testing the filter with a smaller sample to ensure it works as expected.

Security Considerations

The data you input is processed by the selected AI language model. Ensure you are using a trusted and secure language model provider. Review the security policies of the chosen model provider for details on data handling and privacy.