Significant New Magic Cloud Release
For several months we have been working in silence on the largest release we've ever done. Today it was published.
Magic Cloud is what drives your AI chatbot. Some of our clients barely knows that fact. However, Magic has a lot of additional qualities besides being able to create AI chatbots. It's actually a complete Low-Code software development framework.
Among other things it allows for creating backend apps almost like Magic, which is where the name of the platform originates from. If you're not a developer, the word backend might not mean much to you, but an application is divided into two distinct parts.
- The frontend, the stuff you see
- The backend, that is invisible, yet does all the work
Magic Cloud is for backend development.
The point about custom GPTs is that they to a large extent eliminates the need for a frontend, because they provide an alternative to a GUI, being a natural language chat prompt. This allows you to use Magic and OpenAI in combination to create what we refer to as "AI apps".
This release was inspired by OpenAI's release of their GPT platform. A GPT is a custom AI application, anyone can create, allowing you to use ChatGPT and create custom and specialized versions of it. Some examples might be.
- AI apps that will analyze your website
- AI apps that helps you create Facebook ads
- AI apps that functions as your psychologist
The possibilities are endless. However, when you create a custom GPT, you very often want to connect it to your own custom data. We've always had the possibility to connect ChatGPT to your own custom data, and the previous article is actually one of our most popular articles. However, with this release we're increasing the volume 10x in these regards.
Connect your GPTs to your own data
With Magic, and the latest release, you can rapidly create a backend API, and connect it to your GPTs. Watch the following video to understand what I'm talking about.
In the above video I am creating a CRUD API wrapping my database, I am invoking DuckDuckGo to search for queries, scraping websites, and sending emails. More importantly, I am connecting all this functionality to my own custom GPTs in a "drag'n'drop" type of interface, without needing to understand coding or software development.
Low-Code Software Development
If you watch the above video, you will realise that I am creating almost 1,000 lines of code without writing any code at all. This is facilitated for by Magic's Hyperlambda. With Hyperlambda you can create fairly complex applications without coding, which is why its slogan is ...
Where the Machine Creates the Code
The Low-Code parts of Magic have been significantly improved upon in this latest release.
In addition to the above improvements, we've improved upon almost every single aspect of Magic, and it is now much easier and more intuitive to use. Some of these changes are subtle, some of these changes makes it more responsive, while some of these changes makes it more stable by eliminating minor errors and bugs.
- Better documentation - Check!
- Integrated support AI chatbot - Check!
- Using OpenAI as CoPilot to help you create code - Check!
- Better UI and UX - Check!
- 10x better Low-Code and No-Code software development automation - Check!
- Better scraping - Check!
Basically, there's a "bajillion" changes to Magic Cloud and the foundation for our AI chatbot technology, in addition to that Magic as of now allows you to create your own custom GPTs, literally without coding!
We normally try to avoid talking too much about the future, since we try to be "an agile organization". However, I can guarantee you that making it easier to create custom GPTs is a focal area for us. Do you have ideas for a feature you need to be able to create your own custom GPTs? Let us know, and we'll consider it in a future release.
Now that we've got an amazing AI chatbot, focusing on our "creator story" is high on our list of priorities. With AINIRO.IO, AI is "more than just another chatbot", and I suspect the latest release drives home that point.