In the first episode of our saga (if you missed it, check it out here), we laid out the why behind our adventure: a quest to tame the wild, conversation-filled expanse of ChatGPT web interface using the powerful, low-code magic of Mendix. Now, we’re about to delve into the how.
Identifying and Understanding the Stakeholders
In this our adventure we have a cast of characters, but this is something that we must do for every journey we begin. Knowing your stakeholders, their requirements and their needs will be critical for our success.
First off, we have yours truly, the explorer with an array of non-technical needs:
- A way to curate a library of “personas” or prompts, turning a chaotic jumble into an organized masterpiece.
- An efficient system to keep conversations neatly filed away by persona (like a digital librarian).
- The ability to link conversations, because connecting the dots makes for a more meaningful journey.
- And a search function to pinpoint conversations by subject, tag, or title, turning the quest for information into a breeze.
Our second main character is the trusty sidekick, the OpenAI API. Though, in this case, it comes with a comprehensive manual, we might find ourselves interacting with, connecting to, or asking for directions from other third-party in any of our future projects. Be on the ready and get in touch with everyone, understand their capabilities, technologies, connections and most important, key people.
- We need to identify the available models, our supporting cast members in this journey.
- And manage the task of sending and receiving messages from the ChatGPT API, our digital carrier pigeons.
Downloading and Setting Up the Tools
With our cast of characters ready, it’s time to assemble our toolkit for the adventure:
Mendix installation:
We will need a Windows 10 or 11 computer for this one and the rest of the tutorials, but I guess that was an obvious.
Setting up a Mendix account: visit the Mendix website, click on the “Get Started for Free” button, enter your email, and follow the registration process.
Once your account is set, navigate to the “Get Studio Pro” section on the top left menu. It’s like your digital Swiss Army Knife for low-code.
Setting up an OpenAI API account:
- Head over to OpenAI’s portal, sign up
- Create an API key, you can name it anything you may like, but remember to store your API key safely. This is our translator for ChatGPT’s wisdom and we will not be able to see it again for security reasons.
Install Python:
The next two steps are optional, but who would like to go on a journey without a rope and a torch? They are always handy.
Head to Python downloads page here and install the latest version for Windows (if you don’t alreayd have it)
Installing Visual Studio Code:
Visit the Visual Studio website, choose the “Visual Studio Code” , and follow the installation instructions. Even non-coders need a sturdy toolbox, just in case!
Visual Studio Code supports python and OpenAI API has a python module. This will make our API testing calls so much easier to test and implement. Trust me, we will keep the coding part to mostly “copy paste” from OpenAI webpage. People scared of “copy pasting” can stop reading here.
On the Next Episode…
With our characters ready, needs outlined, and tools sharpened, we’re all set for the next stage of our quest.
Since I mostly write for non-technical people, we will keep the journey in small sections, so to not overwhelm anyone. In our next episode we’ll be establishing our first connection with the OpenAI API – our first milestone and a critical component of this adventure (or we will pack and go home).
Join me for the next thrilling instalment of this series, where we’ll get down to the nitty-gritty of forging API connections. Expect deeper insights, and a dash of technical know-how that will make this adventure worth the ride. This is a departure of my usual “low technical” stuff style, but the destination will be worth the effort. Stay tuned, and remember, adventure is out there!
