With Kuika's AI Actions feature, you can add artificial intelligence (Chat GPT-4 and Gemini) as a data source to the applications you develop. You can easily select ready-made templates such as Custom to use in your application and configure them according to your needs. For example, you can integrate CV Reader to read resume data, Create Image to create images, Text to Speech to convert written text into spoken speech, and many more. You can also analyze emotional tones in text with Sentiment Analysis, summarize long texts with Text Summary, and correct grammar errors with Grammar Corrector. In this course, you will learn how to add AI as a data source to your apps using AI Actions and how to use templates effectively.
In order to use AI Actions, you must upgrade your existing subscription plan. By upgrading your subscription, you can benefit from AI-powered automations, smart trade recommendations and advanced AI features. You can visit our Pricing page to upgrade your subscription.
This training content consists of the following topics:
Create an AI connection as a data source.
Smart templates powered by AI.
Create an Action via Artificial Intelligence.
Use Artificial Intelligence actions.
Create an AI Connection as a Data Source
To create a data source connection with artificial intelligence in the Datasources module:
After logging into the Kuika platform, open the project you will work on from the "Apps" screen.
In the "Datasources" (1) module, click on the "+" (2) icon next to Sources on the left side.
3. On the screen that opens, select "AI Actions" (1), enter a name (2) and click the "CREATE" (3) button.
On the screen that appears, you can create a custom AI action or make use of the AI action templates provided.
Smart Templates Powered by AI
After creating AI Action, you can easily perform your operations with the ready-made templates on the screen that opens:
Custom: You can configure it to meet a specific need.
CV Reader: It is used for reading and interpreting the information in resumes.
Create Image: You can create images with artificial intelligence by making explanations according to your needs.
Text To Speech: Converts written text into spoken speech.
Speech To Text: Converts speech to text.
Speech To Text English: Converts English speech to text.
Grammar Corrector: Corrects grammar mistakes in texts.
Taxi Receipt: Used to create or analyze a taxi receipt.
Sentiment Analysis: Analyzes emotional tones (positive, negative, neutral) in text.
Sensitive Content: Detects sensitive or inappropriate content in text.
Text Summary: Summarizes long texts.
Text Extractor: Extracts text from images or PDFs.
Image Analyzer: Analyzes images to provide information about their content.
Article Writer: Creates articles on given topics.
Brand Generator: Creates name or concept suggestions for brands.
Exam Creator: Designs exam questions and answers.
Survey Creator: Creates and configures surveys.
Text Translator: Translates texts into different languages.
PDF Translator: Translates PDF files from one language to another.
Receipt Reader: Reads and analyzes information from receipts.
Receipt Reader Multiple: Reads and analyzes multiple receipts at the same time.
Create an Action via Artificial Intelligence
You can use the Custom template to create a completely unique and customizable function for a specific purpose. After selecting the Custom template, you can make the necessary customizations in the fields below:
1. Action Name: Give this field a name for the action you want to create.
2. System Prompt: Enter instructions, rules, and constraints for the AI.
3. AI Prompt: Define the values and parameters that will be the input for the AI action. You can enter a fixed value or use ‘@’ for dynamic values. For example, you can use the ‘@text’ parameter for texts to be summarized.
4. Template: Select the template you want to use from this field. Kuika offers Generic, Article Writer, Brandar Creator, Exam Creator, Grammar Creator, Personal Trainer, Personal Diet Expert, and Survey Creator templates in AI Action.
5. Model: The models offered by OpenAI are GPT-4, GPT-4 Turbo, GPT-4o, GPT-4o- mini. Gemini-1.5-Flash-001’dir. By default the GPT-4 model is selected.
With Gemini-1.5-Flash-001 you can upload “Image” and “PDF” to your data source.
6. Temperature: Set this setting, which controls the productivity and randomness of the AI, to a value between 0 and 1. 0 produces more consistent responses and 1 produces more random responses.
7. Maximum Length: Set the maximum number of characters, words or tokens that can be generated. This controls the length of the response to be generated.
8. Stop Sequences: Enter the stop word or sequence of words in this field to make the response stop at a specific point.
9. Top P: Achieve more consistent and contextually appropriate output by selecting the most likely words that pass a certain probability threshold (p). Enter a value between 0 and 1; for example, selecting 0.5 considers half of all probability-weighted options.
10. Frequency Penalty: Use this setting to reduce the probability of using repetitive words or lines. Enter a value between 0 and 1.
11. Presence Penalty: Make it easier to create new content by adding various tokens to the text. Enter a value between 0 and 1.
Finally, create your action by clicking the "CREATE" button in the upper right corner.
Use Artificial Intelligence Actions
You can connect the artificial intelligence actions you created to your screen designs in the "UI Design" module.
To connect actions;
Open the "UI Design" module. Go to the screen where you want to add the action. You can connect the action to the screen or elements according to your usage scenario.
2. Connect the artificial intelligence action "AI Actions" in the Custom category via "+ADD ACTION" in the Properties panel.
