Skip to content

This example adds AI-powered filter capabilities to the DevExpress VCL Data Grid (using a lightweight ONNX model and a BERT tokenizer).

License

Notifications You must be signed in to change notification settings

DevExpress-Examples/vcl-grid-smart-semantic-filter

Repository files navigation

DevExpress VCL Data Grid - Smart Semantic AI Filter

This example adds AI-powered filter capabilities to the DevExpress VCL Data Grid (using a lightweight ONNX model and a BERT tokenizer). Unlike built-in search/filter operations, Smart AI Filter compares strings based on semantic similarities rather than exact matches.

An application that displays a search query "accessories" and a list of items filtered using the AI-powered smart search

Third-Party Dependencies

Prerequisites

  • Microsoft Windows 10 (64-bit) or newer
  • ONNX Runtime v1.23.0 (required to run on Windows 10)
  • Embarcadero RAD Studio IDE 12 or newer (Community Edition is not supported)
  • DevExpress VCL Components v25.1.5 or newer

Deploy and Run

  1. Clone this repository (including submodules):

    git clone --recurse-submodules git@github.com:DevExpress-Examples/vcl-grid-smart-semantic-filter.git

    If you have already cloned the repository without submodules, initialize them using the following command:

    git submodule update --init --recursive
  2. Open and build the SimilaritySearch project (DPR).

  3. If using Windows 10 (64-bit), place the latest version of onnxruntime.dll (available in the official repository) into the executable file folder.

  4. Run the sample project.

Test the Example

Type a word or phrase into the Smart Search AI Filter field and click Apply.

Does this example address your development requirements/objectives?

(you will be redirected to DevExpress.com to submit your response)

About

This example adds AI-powered filter capabilities to the DevExpress VCL Data Grid (using a lightweight ONNX model and a BERT tokenizer).

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 5

Languages