Creating agents in Tess is a strategic option for applying Artificial Intelligence to maximize productivity and scale daily activities. In this article, we will explore step by step how to develop agents that not only meet user needs but also are optimized to ensure better visibility and usability.
One of the most common challenges when creating your templates in Tess AI is crafting titles and descriptions that are clear and informative. Often, we come across templates with repetitive or ambiguous descriptions that can end up confusing users, preventing them from fully understanding the functionality and purpose of the template in question.
To overcome this obstacle, it is recommended to use the auxiliary template created by our CEO, Rica Barros, called "Automatic Titles and Descriptions," which is available in Tess AI. By adopting this practice, you will not only improve the usability of your templates but also enhance the efficiency of communicating their functionalities to end users.
Utilizing advanced steps in agents in Tess AI can mean that you are further enhancing and training your AI, expanding its usage and capability to perform distinct actions, especially when it comes to manipulating and extracting data from complex formats like PDFs and URLs.
To facilitate understanding and adoption of these advanced features, we will simulate the step-by-step process of how to set up and use the step to attach a PDF.
Thus, we summarize the process as follows:
- Create the input field with the option to 'Upload a file';
- In the AI Steps option, add the step 'File|Read PDF';
- In the step 'File|Read PDF', the option 'Insert the file or link to the PDF' must be filled with the variable from the input field created to attach the file.
This process is exemplified in the screenshot below:
[IMG]
Input fields are spaces that will bring versatility to the application of your template. That is, they are the information or objectives that a user may specify before running your template, which will affect the final outcome.
To ensure that these users understand what needs to be filled out, we recommend being very clear and direct when naming the input field, its description, and tips/examples of input.
The example below depicts a good case of this recommendation:
[IMG]
Note that the objective of the field has been named, there is an explanation in the description, and a helpful example in the tooltip. This will help more people understand your template and know how to use it effectively!
Additionally, it is not advisable to create input fields that define the maximum number of words in the final result and the language. To address this type of configuration, there is an option available at the bottom of the page (screenshot below). If you enable this option for the user, they will also be able to define the type of creativity and the AI model.
[IMG]
This is an option that provides a more controlled experience, ensuring that the final result aligns with the expectations of those who created the template.
Creating prompts does not have a fixed rule; however, it is necessary to ensure that it aligns with the objective of the template. Thus, it is recommended to structure prompts clearly and objectively, as well as to test different versions, more or less detailed, so you can refine them as many times as necessary.
With these best practices, we hope you can successfully create various templates for use in personal and professional routines.
If you need further assistance, you can contact our support!
