ChatGPT – Personal Assistant with Power Apps

ChatGPT can be used in a variety of ways, including Natural language processing, this means that it can be used to understand and generate human-like responses to text-based queries, allowing for more natural and efficient communication between humans and computers.

This gives an idea of what approach we could take to solve the problem of having a personal asistant!

A personal assistant is a professional who provides administrative and personal support to an individual or a group of individuals. They typically manage tasks such as scheduling appointments, arranging travel, managing emails, conducting research, making phone calls, and handling other administrative tasks that the individual or group they work for may need assistance with.

In recent years, advances in technology have led to the development of virtual personal assistants, which are digital assistants that can perform similar tasks as human personal assistants, but without the need for a physical presence. These virtual assistants are typically powered by artificial intelligence (AI), to be precise – in this era – ChatGPT-4.


OpenAI allows us to interact with the API through HTTP requests from any language and uses API keys for authentication.

API Reference – OpenAI API

API Reference – OpenAI API

To create a API Keys, go to your OpenAI account and Create a new secret key

OpenAI – API Keys

Power Platform – Custom Connector

In Power Platform, a custom connector is a tool that allows you to create a connection between your app and an external service or system. With a custom connector, you can extend the functionality of Power Apps by integrating with APIs and services that are not natively supported.

By following these steps, create a connector that is tailored to our specific needs and requirements:

Navigate to the Power Apps portal and select “Custom connectors” from the left-hand menu, then Click on the “New custom connector” button to create a new connector.

1. Under General tab: Fill in the basic information about your connector, such as the name, description, and logo – in the Host textbox, enter:

2. Under Security tab: fill as follows

3. Under Definition tab:

3.1 Click on New action – Define the actions and triggers that your connector will support. An action is an operation that your connector can perform, such as creating a new record or updating an existing one. A trigger is an event that your connector can respond to, such as when a new record is created or updated.

3.2 Click on Request and fill as follows (as per the OpenAI Documentation), then click Import

API Reference – OpenAI API

Verb: POST


Headers: Content-Type: application/json

Body: { “model”: “gpt-3.5-turbo”, “messages”: [{“role”: “”, “content”: “”}]}

3.3 Request > edit the Content-type property > provide a default value and make it required and internal (see capture)

3.4 Request > edit the Body property > make it required and important (see capture)

repeat the same for Body payload property (see in green) – content, role, model – edit each of them and make it required and important.

4. Skip the Code (Preview) tab.

5. Under Test tab:

5.1 Click New connection, add your API Key which you created (see above) in the following format: Bearer sk… then click Create

Then fill as follows and click on Test operation

5.2 Copy the Response Body and set it as a Default Response in the Definition (see next step)

5.3 Under Definition tab, click Add default Response and paste the Body Response then click Import, then Update the custom connector.

Integrate the connector with Power Apps

1. In Power Apps, create a Blank Canvas App, insert a Text Input and a Send Icon as follows

2. Under Data, add the ChatGPT Connector

3. On the Send icon, OnSelect property – set the following code to call the the ChatGPT API via the custom connector

Set(varChatGPTAPIResponse, ChatGPTConnector.ChatGPTApi(“gpt-3.5-turbo”, Collect(collectionChatMessages, {role:”user”, content:TextInput1.Text})));

the variable varChatGPTAPIResponse contains the role and content properties which you can easily manipulate as per your convenience (Gallery, Datatable, etc).

Voila! this is how our personal assistant looks!

Visual ChatGPT

Visual ChatGPT is a version of the ChatGPT language model that has been trained to generate text-based responses to visual prompts. It uses computer vision algorithms to analyze and interpret images, and then generates text based on the content of the image.

To get started with Visual ChatGPT, you will need to provide it with an image that you want to generate a response for. This can be done by either uploading an image file or providing a URL to an image hosted online.

Once the image has been processed by the computer vision algorithms, Visual ChatGPT will generate a response in natural language based on the content of the image. This response can be used to answer questions, provide information, or generate captions for images.

To use Visual ChatGPT, you can interact with it through a chat interface, similar to how you would interact with a human chatbot. Simply provide the image and wait for Visual ChatGPT to generate a response. You can also provide additional context or information to help guide the response generated by the model.

Enable Visual ChatGPT on your Windows Machine

Folow the steps:

# clone the repo
git clone

# Go to directory
cd visual-chatgpt

# download Ananconda & create a new environment

conda create -n visgpt python=3.8

# activate the new environment
conda activate visgpt

#  prepare the basic environments
pip install -r requirements.txt

# Generate an API Key/Secret Key from your 
(see image below)

# set your private OpenAI key (for Windows)
set OPENAI_API_KEY={Your_Private_Openai_Key}
set OPENAI_API_KEY=sk-...

# Start Visual ChatGPT !
# You can specify the GPU/CPU assignment by "--load", the parameter indicates which 
# Visual Foundation Model to use and where it will be loaded to
# The model and device are separated by underline '_', the different models are separated by comma ','
# The available Visual Foundation Models can be found in the following table
# For example, if you want to load ImageCaptioning to cpu and Text2Image to cuda:0
# You can use: "ImageCaptioning_cpu,Text2Image_cuda:0"

# Advice for CPU Users
python --load ImageCaptioning_cpu,Text2Image_cpu

# Advice for 1 Tesla T4 15GB  (Google Colab)                       
python --load "ImageCaptioning_cuda:0,Text2Image_cuda:0"
# Advice for 4 Tesla V100 32GB                            
python --load "ImageCaptioning_cuda:0,ImageEditing_cuda:0,
OpenAI – API Keys


The Visual ChatGPT demo should now be running on your local machine. You can open a web browser and navigate to http://localhost:7868 to interact with the model.

It’s important to note that while Visual ChatGPT can generate responses based on visual prompts, it is not perfect and may sometimes generate inaccurate or inappropriate responses. As with any AI model, it is important to use it responsibly and carefully evaluate the accuracy of its responses.