In the past six to nine months we have all witnessed the rise of AI. New tools or existing software vendors that include AI functionality seem to pop up every day. Most of us have used chatgpt by now for research or tasks like copywriting and editing. In programming, AI tools have been used for writing code, quality assurance and building small applications. They have significantly enhanced the speed of application development. At Nymbl we have used AI tools since the beginning of 2023 and have seen a tremendous improvement in productivity.
And it is not only us:
According to a report by Gartner by 2027, 70% of professional developers will use AI-powered coding tools, up from less than 10% today. (Gartner Article)
Substantial gains in development speed have been realized by many. Github quoted quantitative research on their blog for example which stated that developers using Github Copilot coded up to 55% faster (link to article)
No code and low code application platforms have also jumped on the hype train and implemented AI capabilities into their feature stack. The question this article wants to answer is then, how do no code and low code platforms incorporate AI and how does this benefit the developers?
We will not cover building AI enhanced solutions, this will be an article for a later time. Instead we focus on AI enhanced development and dive deeper into the question above as well as share our research on three widely used and industry leading platforms.
How did Mendix integrate AI into their platform?
Mendix is providing AI-assisted development bots called MxAssist. It is targeted at Citizen developers to provide guidance and ensure best practice is used. Moreover it helps experienced developers to remove inefficiencies and increase overall performance
It has 3 Main functions:
1. The Mendix Logic Bot provides context driven recommendations on how to build workflows (called microflows in Mendix). It helps developers build microflows 30% faster and is based on Machine Learning from more than 12 a million anonymized application logics built with Mendix. It provides real-time recommendations for next actions with claimed 95% accuracy and streamlines development by pre-populating paraments among other features.
2. Next up is the Mendix Best Practices Bot. It automatically inspects and benchmarks app models against Mendix best practices, can detect anti-patterns and automatically refactors and app models to use best practices and increase performance.
3. Lastly Mx Assists Data Validation Bot automates manual and repetitive tasks of data validation. It can generate a validation microflow simply by right-clicking on a supported element for example. Additionally it automates the building of specific elements on there pages where a data validation might be required later on.
”Mx Assist is really a game changer for citizen developers. The most effective area where MxAssist helps out is logic flows, aka microflows, where it suggests the next best action to take, according to how it looks already. Especially less experienced developers can learn well from it, understanding what it is suggesting and why it is suggesting it.” Says Tyler Stubbs, Technology Partner for Mendix at Nymbl.
How did Flutterflow integrate AI into their platform?
Flutterflow is a mobile app builder based on Google’s flutter software development kit. They recently released their Flutterflow AI Gen. It is an AI assistant that supports the whole app development journey from design to build and launch.
AI Page Gen provides instant page to text. Developers can use written prompts and the AI page gen builds a first draft that can then be customized and also exported as flutter code.
AI Schema Gen creates schemas from written prompts and transforms them into your backend schema
AI Theme Gen uses, you guessed it, text prompts to create beautiful and appealing color schemes. This removes the need for manual hex code matching for example and makes design so much easier.
Lastly the AI Code Gen is your AI Co-pilot in the browser. It provides code suggestions from prompts which are instantly usable.
Overall Flutterflows AI Gen looks like a chatgpt for mobile app building and makes development even faster and easier than no code already does. At Nymbl we envision this will enable even more builders to build digital products and mobile process automations tools.
How did Webflow integrate AI into their platform?
As of this writing Webflow is deeply investing into AI as per a blog post from their Co-Founder & CEO Vlad Magdalin. He believes that ”the tools that let you deeply customize, manipulate, and build upon what AI generates will ultimately create the biggest impact”.
So what is Webflow working on?
Webflow is a very powerful visual development platform with lots of features. Extensive help sites and the Webflow University exist to help guide users and speed up learning. However users often need to leave the platform and change to another window or tab in their browser. So Webflow wants to simplify this by integrating prompt based AI support into their solution and leveraging all the useful content out there on Webflow.
Secondly Webflow is working on their AI Co-Pilot. Many features are planned. Some of the prominent benefits will be:
- Quickly generate, alter, or summarize static and CMS content with simple prompts
- Auto-generate optimized SEO settings based on the actual content of the page
- Auto-generate image alt tags to improve accessibility
- Create unique new images in the Assets panel based on a text prompt
- Generate code in the Custom Code editor based on a prompt
- Quickly translate content into different languages
For beginners of a no code or low code platform there are only few things as daunting as starting with an empty template.
So lastly Webflow wants to make it easier to create the first templates and get started by using prompts and a simple builder based on that. You can find the super straightforward user interface in the picture below
What is next and what is our experience with AI assisted no code and low code development?
In this article we mention three no code and low code platforms specifically in Mendix, Flutterflow and Webflow. But there are many more platforms that offer these capabilities and we are just scratching the surface. The no code and low code vendors that do not offer it yet, are announcing it, like Bubble did during their Bubblecon conference this year. You can see their discussion under this link: here
Application development has already increased in speed just from implementing these first versions of AI assistants. As more and more applications are being built on these platforms each day using AI assistants they will get more and more powerful and accurate over time. While now there might still be re-work and adjustments necessary these will become less and less.
We estimate that in a year's time all leading no code platforms will integrate AI assistants and that development speed will double or triple from this alone as iterations are faster and natural language can be used. Moreover these assistants will make it easier for less experienced citizen developers to pick up no code and low code tools and therefore enhance the developer pool and quality of applications built even more.
”AI enables development teams to execute faster, either through a platform or through the generation of quality code (or no code). Additionally It supports requirements gathering. For example AI assistants can speed up the context and detail of writing user stories. We are working to include more and more of this functionality to provide the best state of the art application development and technology agnostic advisory services to our clients” Says Nymbl Founder and CEO, Martyn Mason.
We hope this article gave you a first taste of the topic and generated some insights for you. If you have any questions about AI capabilities of these tools or other in the market or want to integrate AI functionality into your mobile or web applications, contact us by clicking here.