惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

B
Blog RSS Feed
量子位
Y
Y Combinator Blog
大猫的无限游戏
大猫的无限游戏
B
Blog
U
Unit 42
C
Check Point Blog
I
InfoQ
aimingoo的专栏
aimingoo的专栏
雷峰网
雷峰网
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
博客园 - 【当耐特】
人人都是产品经理
人人都是产品经理
The Cloudflare Blog
H
Help Net Security
MongoDB | Blog
MongoDB | Blog
博客园 - Franky
H
Hackread – Cybersecurity News, Data Breaches, AI and More
J
Java Code Geeks
Microsoft Azure Blog
Microsoft Azure Blog
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
云风的 BLOG
云风的 BLOG
宝玉的分享
宝玉的分享
爱范儿
爱范儿

Mendix

Mendix Release 11.14 | Mendix Why Domain Experts Are Critical to Building the Agentic Enterprise | Mendix From Workflows to Agentic Orchestration  | Mendix Mendix Release 11.13 | Mendix Mendix Portable Runtime: Deployment, Simplified | Mendix Computable Awards 2026: Mendix Is Nominated! | Mendix Mendix 11.12 LTS: Strategic AI for Enterprise Leaders | Mendix Mendix Release 11.12 | Mendix Intelligence Center X Closes the AI Value Gap in Industrial | Mendix Mendix Release 11.11 | Mendix Faster LTS Releases: Your Strategic Upgrade Roadmap | Mendix Mendix Release 11.10 | Mendix Bringing Teamcenter BOMs to Every Corner of Your Mendix Apps  | Mendix Mendix Release 11.9 | Mendix Dynamic Reporting Using Mendix GenAI | Mendix The Right Model for the Job: A Developer's Guide to Choosing LLMs in Mendix | Mendix Mendix Release 11.8 | Mendix Application Development in the Enterprise: Why AI Needs Architectural Guardrails | Mendix Vibe Coding Mendix Widgets  | Mendix Exploring Agentic AI in Industrial Manufacturing | Mendix Release 11.7 - Zero Downtime, AI Tools & React Ready | Mendix How to use GenAI in Industrial Manufacturing | Mendix How can you use AI in manufacturing? We asked an expert. | Mendix Beyond Chatbots: Building Agentic Apps with Altair & Mendix | Mendix Introducing Mendix Cloud Private Connectivity (Public Beta) | Mendix Release 11.6 - Zero Downtime Deployments & Workflow Advances | Mendix Celebrating 43 New Mendix MVPs in 2025 | Mendix Mendix on Azure Goes GA: Secure, Proven, Scalable | Mendix Release 11.5 - Smarter Flows, Stronger Mobile & Self-Service AI | Mendix 5 Workflow Features You Might Not Know Exist | Mendix
What If AI Closed the Gap Between Planning and Building? ...
Marek Jednaszewski · 2026-05-06 · via Mendix

AI-assisted development has come a long way, fast. We went from autocomplete to code generation, and the next shift is already here: agentic development. With Maia, we’ve built that capability into the core of the Mendix Platform — not as a bolted-on feature, but as a fundamental change in how apps get built, from the first rough idea all the way through to working application code in Studio Pro.

What is agentic development?

Agentic development changes the relationship between developer and AI. Instead of responding to precise commands, the AI takes initiative — planning, generating, and refining work based on a goal, while keeping you in control of what moves forward.

The key distinction is scope. A traditional AI assistant answers questions or generates snippets when asked. An agentic one understands what you’re trying to build, sequences the work, surfaces decisions that need your judgment, and executes across multiple steps without requiring you to micromanage each one.

That’s not automation without oversight. It’s intelligent execution under human direction — and it’s what Maia is built to do.

Maia Plan – AI-powered planning

Great applications start with clarity on intent. And that clarity matters most at the very beginning.

The earlier teams align around goals, scope, and requirements, the cheaper and easier change is. But as development progresses, misalignment becomes expensive — leading to rework, delays, and friction between team members. Teams spend a significant portion of their delivery effort, often up to 25%, on requirements clarification, refinement sessions, and rework caused by ambiguity.

Yet most projects don’t begin with clarity. They begin with scattered notes, stakeholder conversations, slide decks, and half-defined ideas.

Maia Plan is the AI-powered planning engine within the Mendix Portal that transforms those messy inputs — workshop notes, project briefs, intake discussions, or even a rough prompt — into structured, high-quality, execution-ready delivery plans. Upload a single document or multiple files to give Maia the full context it needs. The result: a build-ready backlog in hours, not days.

From chaos to clarity: how Maia Plan works

Maia Plan first creates a clear project scope with defined goals, measurable success criteria, target users and personas, functional and non-functional requirements, and explicit scope boundaries. This establishes alignment early, ensuring everyone understands what is being built and why — before a single microflow gets created.

From there, it generates a structured solution proposal: well-defined epics, user stories optimized for Mendix development, acceptance criteria, technical recommendations, and logical sequencing. Maia also surfaces relevant components from the Mendix Marketplace — including your organization’s private Marketplace — based on the stories it generates, so teams know what already exists and can reuse proven solutions rather than build from scratch. If your organization uses the Company-Approved setting, only internally vetted components get recommended.

Your team reviews and adjusts everything through a conversational interface before anything is approved. Human judgment stays at the center of every decision.

Once approved, the artifacts land directly in Epics — no manual story-writing, no copy-paste. From there, stories flow into Maia Make in Studio Pro with a single click, where pages and logic get scaffolded based on the agreed plan. A working first version can be in stakeholders’ hands for feedback within minutes of development starting.

This isn’t a return to waterfall. Maia Plan doesn’t lock teams into heavy upfront documentation or rigid specifications. It brings the best of both worlds: the clarity and structure traditionally associated with detailed planning, combined with the speed and continuous validation of agile development. Structured intent early, fast iteration throughout.

No handovers. No translation gaps between what the business asked for and what gets built. Less rework. Faster delivery.

Maia Make – one conversational interface for everything in Studio Pro

Maia Make is where the plan becomes the product. Once your backlog is approved in Maia Plan, stories flow directly into Maia Make in Studio Pro — no handover meeting, no translation layer. Select the stories you want to tackle, and Maia starts building.

Maia Make brings all of Mendix’s generative AI capabilities into a single unified chat interface inside Studio Pro. Describe what you need, and Maia generates the necessary application components — data structures, pages, microflows, and more — all in one continuous conversation, without breaking your flow.

Under the hood, Maia Make runs on Claude Sonnet 4.5, Anthropic’s model optimized for Mendix’s modeling environment.

What Maia Make can build

Describe your data model and Maia creates the entities, attributes, and relationships that underpin your app. Describe a screen and Maia builds the page — or better yet, drop in a screenshot, wireframe, or design mock-up and Maia replicates the layout directly. Describe a business process and Maia generates the microflow logic, including conditions, expressions, and data operations.

Beyond the core building blocks, Maia can also generate constants, JavaScript actions, and set up folder structures — the kind of setup work that’s necessary but rarely the best use of a developer’s time.

One capability worth calling out specifically: Maia Make works with MCP (Model Context Protocol) in both directions. As an MCP client, it connects to external tools — pull in design assets from Figma, query external systems, or bring in context from whatever tooling your team already runs. As an MCP server, Studio Pro exposes its own generation tools to the outside — meaning your own AI agents or automation pipelines can connect to Studio Pro and trigger microflow generation, page creation, domain model generation, and more, directly from whatever agent or tool you’re working with.

Throughout all of it, Maia keeps you in the loop. It explains your microflows and pages in plain language on request, shows its work, and waits for your review before changes are applied. You stay in the driver’s seat.

From idea to application — without the gaps

Maia Plan and Maia Make are designed as two halves of the same loop. Maia Plan turns scattered inputs into a structured, agreed-upon backlog — with Marketplace component recommendations built in so teams start development knowing what to reuse. Maia Make takes that backlog into Studio Pro and turns it into a working application through a single conversational interface.

The design principle across both is consistent: you stay in control. Maia suggests, generates, and executes, but your review and approval drive what moves forward at every step. It handles the scaffolding so you can focus on the decisions that actually require your expertise — the business logic, the edge cases, the things that make your application yours.

Try Maia and see where it takes you

What would your last project have looked like with Maia? Find out — try it today. And if you’re on an older Mendix version, the 11.12 LTS release in June is your moment to get current and stay there.