print

Mendix AI-Assisted Development

How Does Mendix Leverage AI to Help Users to Build Applications?

Mendix leverages Artificial Intelligence (AI) and Machine Learning (ML) to help development teams model and deliver Mendix applications faster, with more consistency and with higher quality. This is an emerging trend in software development, commonly known as AI-Assisted Development (AIAD). AIAD in the Mendix platform is called Mendix AI Assistance (Maia). Maia consists of different capabilities that act as virtual co-developer capabilities, providing guidance, assistance and generation in a certain domain or stage of the application lifecycle development. Currently, Maia consists of several virtual co-developer capabilities: In Studio Pro, we have Maia Chat for developer guidance, Maia Logic and Workflow Recommenders, Best Practice Recommender for in-editor assistance, and Generative AI based features like Translation Generator. Next to this, we provide Maia Rewrite and Summarize on the Mendix Community.

How Does Mendix Leverage AI to Help Users to Build Application Logic Faster with Higher Quality?

Maia Chat guides developers during their work in Mendix Studio Pro. Developers can ask questions about app development in Mendix, including how to apply concepts, best practices, and development patterns. Maia Chat uses Generative AI based on large language models (LLMs) trained on Mendix Documentation, Community posts and Academy learning paths. It is tailored specifically to Mendix questions and provides relevant answers and references to original documents. For more information on how to use Maia Chat, see the Maia Chat documentation in the Mendix Studio Pro Guide.

Mendix enables you to visually build application logic easily with microflows, nanoflows and workflows instead of having to write code. To make this ever easier, the Maia recommenders for all three logic editors (which provides AI-powered suggestions) guide you through modeling and configuring your application logic (microflows, nanoflows, and workflows). It gives you a real-time and context-driven list of the next best actions based on the application logic already designed and other context-related information. The Maia recommenders enhance developer productivity and learnability by suggesting the next best actions in developing application logic.

The Maia Logic Recommender is built using machine learning analysis of over twelve million anonymized application logics (microflows)—built with Mendix over a decade—to detect and learn the best practice patterns in microflows.

The key features of Maia recommenders are the following:

  • In-editor next best action suggestion – recommends the top 7 next best parameterize actions
  • Contextual suggestions – derives context in different ways, including “looking” left and right in a logic when the developer inserts a new element or action mid-flow and inferring the context using the page where the logic is used
  • Search-based suggestions – quicky finds any parameterized action developers need
  • Auto-configuration – not only provides the next best action, but also automates further development by pre-populating the parameters for such an action
  • Combined with mouse and keyboard navigation, the recommenders provide the best way for new developers to learn the next best action in any context and unrivaled development speed for the advanced developers

For more information on how to use Maia recommenders, see the Logic Recommender and Workflow Recommender documentation in the Mendix Studio Pro Guide.

How Does Mendix Leverage AI to Help Users to Build Applications According to Mendix Best Practices?

Development teams often spend substantial time training, enforcing, and peer-reviewing the implementation of best practices. Even then, new developers might follow some anti-patterns that are hard to detect during development and fix after deployment.

Maia Best Practice Recommender is an intelligent virtual co-developer capability that helps you improve your app by inspecting your app model against Mendix development best practice in Mendix Studio Pro. It detects anti-patterns during the design and development, pinpoints you to these anti-patterns, suggests you how to resolve it, and in some cases can automatically fix these issues.

It consists of a three-level assistance:

  1. Detection – it inspects the model, identifies issue, and pinpoints you to the document/element causing the issue.
  2. Recommendation –  it explains the identified issue, the potential impact, and how to fix it. There is also a detailed best practice guide with a dedicated step-by-step guideline of how to fix the issue.
  3. Auto-fixing – it can automatically implement the best practice and fix the issue.

 

Maia Best Practice Recommender is built using statistical analysis of thousands of anonymized Mendix app to learn common anti-patterns as well as using Mendix Expert Services best practices in the development of microflows, domain models, pages, security, and so on. Maia Best Practice Recommender enhances development efficiency by substantially reducing peer reviews, educates junior developers on best practices, increases developer productivity via the automatic detection and pinpointing of issues, and provides assistance for addressing these issues. For more information on how to use Maia Best Practice Recommender, see the Best Practice Recommender documentation in the Mendix Studio Pro Guide.