PlantCare is an example of ecommerce based on an online store for selling outdoor and indoor plants. It combines two key components in the development of modern solutions: a Backoffice for managing administrative functions, and a customer-facing Frontend that provides an intuitive and engaging user experience.
This solution was developed using GeneXus Next, which accelerates development times and facilitates the integration of advanced capabilities through the use of generative artificial intelligence.
All generative artificial intelligence features provided by GeneXus Next are automatically configured to run on the Globant Enterprise AI platform, which enables the necessary platform scaling and ensures independence from LLMs, regions, clouds, and cloud providers.
Both the Backoffice and the Frontend rely on Agent objects that streamline tasks and unlock the full potential of the application.
The Frontend is generated with Panel objects, from which GeneXus generates progressive web applications (Angular Frontend) that can be run in device browsers or on any computer or device that runs a browser.
The Backoffice enables employees to efficiently manage key business information through features such as content insertion, editing, and deletion (CRUD operations). In addition, thanks to the use of intelligent agents, data entry becomes much faster and easier.
You can try it out directly here: https://samples-next.genexus.com/PlantCare/home
- Artificial intelligence agents for data loading: Enter the name of a plant. Based on that input, the agent completes the related information by returning a formal description of the plant, subtitle, irrigation and a category selected from those already stored in the database.
The Frontend follows Figma patterns and aims to be intuitive, visually clear, adaptable, and enhanced by artificial intelligence.
You can try it out directly here: https://apps-angular.genexus.com/PlantCare/EHome
- Semantic search: You can explore plants and select various options of interest. The results will display the plants that best match the request based on similar concepts rather than exact matches.
- Agentic and Conversational UX: Leveraging the capabilities of the Conversational Chat Control, the interface allows you to make a wide range of queries related to the plant environment. The Chat control is connected to an Agent that retrieves data from both the corporate database and public internet sources, merging the information into the UX and UI in a conversational way.
- GeneXus Next Desktop
- Apache License 2.0
1. Get GeneXus Next Desktop.
2. On the Start Page, the "PlantCare" sample is available. Select it.
3. Select Build > Rebuild All in the GeneXus main menu.
4. Select View > Launchpad in the GeneXus main menu.
- To run the Backoffice: Click on "Home"
- To run the Frontend: Click on the "EHome" Panel
Sample data load
The application includes an integrated test dataset for testing it after the build process has been completed.
To this end, the “initPlantCare” program will be used. It is a GeneXus main Procedure object that can be executed by command line.
When you run it, the following output will be displayed:
========== Execution Started ==========
dotnet "/app/kbs/PlantCare_Sample/Cloud.NET/web/bin/batch.ainitplantcare.dll"
** Initialize PlantCare Data **
Data successfully unzipped.
1. Importing Product: Ficus Lyarata
2. Importing Product: Asplenium
3. Importing Product: Zamioculca Zamiifolia
.......
41. Importing Product: Coleus
Data successfully imported.
Success: Execution
Success: Run initPlantCare
Several features and criteria of the development process are highlighted below.
Several design components can be reused. Different Stencil objects have been created, enabling the reuse of design components and thus increasing productivity.
In some cases, when the logic can also be reused or encapsulated, Components were used.
In the Backoffice, an Agent object was created to assign default values to the attributes of the Product Transaction.
In the Frontend, the Chat Control was used to assist users in choosing the most appropriate product. The Chat control is connected to an Agent that retrieves data from both the corporate database and public internet sources, merging the information into the UX and UI in a conversational way.
In addition, the search page sorts results based on embedding proximity, allowing users to find products of interest using natural language.
To simulate a shopping scenario, a session per device is created using the ClientInformation.Id property. The transactions associated with this functionality are Session, SessionFavorite, and SessionCart.
To turn this example into a real application, you will probably need to use GeneXus Access Manager (GAM) for security and user management.
In addition, because the checkout function is not implemented, you can extend the example to support online purchases by using APIs to integrate with payment methods.
The Header Panel object is componentized to encapsulate its operation. This allows its events and characteristics to be programmed in one place and reused across most Panels. A relevant feature is the use of the GlobalEvents external object, which is used to update the cart items.
Get support through the Issue Tracking system.