Since then, more of our team members have gained experience with OutSystems, prompting us to conduct another trend analysis to explore the future outlook of low-code tools.
What Are Low-Code and No-Code?
The adoption of low-code and no-code tools is increasing to improve app development efficiency, enabling faster development and cost reduction.
Low-Code: Requires minimal coding for app development. Some programming knowledge is necessary, but it allows for flexible customization.
No-Code: Enables intuitive app creation without coding. However, customization options are more limited.
[Comparison table with representative tools]
Comparison of Major Low-Code & No-Code PlatformsThe market share of low-code and no-code platforms is primarily based on industry recognition and the number of companies adopting them. Below are the leading platforms ranked by market share:
Microsoft Power Apps
A widely adopted platform due to its strong integration with Microsoft products (especially Office 365 and Azure). As part of the Microsoft ecosystem, it allows businesses to easily create and manage applications.
Adoption: Used by organizations of all sizes worldwide, from small businesses to large enterprises.
OutSystems
A low-code platform specializing in enterprise solutions, excelling in large-scale and complex app development. It features strong automation and AI functionalities and is particularly popular among large corporations.
Adoption: Commonly used in industries like finance, manufacturing, and public institutions.
Mendix
A low-code platform designed for enterprise applications, known for its integration with SAP and industry-specific approach. It supports both cloud-based and on-premises deployment, providing an advanced development environment.
Adoption: Frequently used in manufacturing and logistics industries.
Appian
Specialized in Business Process Management (BPM), this platform excels in process automation and complex workflow management. It is designed for enterprises looking to optimize business operations.
Adoption: Popular in industries such as banking, insurance, and pharmaceuticals.
Future of Low-Code and No-Code
The future of low-code and no-code tools looks promising, with market growth and technological advancements driving increased adoption. The integration of AI will further enhance development speed and expand possibilities for both enterprises and individual developers.
Market Growth
The global low-code/no-code market was valued at $17 billion in 2022.
It is projected to grow to $45 billion by 2027.
The market is expected to expand rapidly over the next 5–10 years.
Key Trends in the Low-Code/No-Code Market
Stronger Integration Between AI & Low-Code/No-Code AI will automate development processes, allowing for natural language-based app development, eliminating the need for manual coding. Additionally, AI will optimize UI/UX design and code generation.
Examples of AI Integration:
Power Apps’ “Copilot” and OutSystems’ “AI Mentor”
Developers can input commands like “Create a product management app”, and AI will automatically generate the application.
AI will analyze user behavior and optimize UI layout automatically.
With further AI advancements, even complex logic and data analysis will be handled by AI, accelerating the shift toward true no-code development.
Developers’ roles will transition from coding to designing and managing applications.
Conclusion
Based on market share, Microsoft Power Apps dominates the low-code space due to its strong integration with Microsoft’s ecosystem. However, OutSystems also maintains a strong position, particularly in enterprise applications.
While large enterprises have been the primary adopters, the question remains whether low-code tools will expand further into SMEs and general users.
Looking ahead, as low-code tools and AI continue to merge, developers will focus more on designing applications, while AI handles coding. Exploring how OutSystems can integrate with AI will be a key consideration moving forward.
For our next post, we’ll explore the potential of OutSystems and AI integration and what it can achieve. Stay tuned!
In this engirring blog, we introduce the learning topics we work on daily.
What is Local-First?
Recently, the concept of Local-First has been gaining attention. Many applications are designed with an online-first approach, where data is stored in the cloud. This allows users to access their data from any device and easily share it with others. However, applications that rely on online connectivity have several disadvantages, such as:
Data is inaccessible without an internet connection.
The application becomes unusable if the server goes down.
Communication with the server can introduce delays, reducing responsiveness.
Storing personal data in the cloud poses security risks from external access.
The Local-First approach addresses these issues. In Local-First applications, data is primarily stored on the user’s device and synchronized only when needed. This approach offers several benefits:
Data remains accessible even without an internet connection.
The application functions independently, unaffected by service outages.
Immediate data read/write operations without relying on a server.
Personal data is managed locally, reducing dependency on the cloud.
Examples of Local-First Applications
Several applications utilize the Local-First approach, including:
Evernote: Allows users to create and view notes even while offline. Synchronization with the cloud ensures data availability across devices.
Notion: An all-in-one productivity tool featuring document management, task tracking, and database functionality. Users can edit content offline, and changes are synced to the cloud to maintain data consistency.
Let’s Build an Application! (TypeScript Edition)
To experience the Local-First approach, let’s build a simple To-Do App that runs entirely within a browser! This application will store data locally, ensuring that tasks remain saved even after a page reload.
Below, we introduce the key implementation details.
ToDo の追加（データをローカルに保存）
async function addTodo(text: string) {
const todo = {
_id: new Date().toISOString(), // 一意のID
text,
completed: false
};
// データをローカルに保存
await db.put(todo);
}
ToDo の表示（保存されたデータを取得）
async function renderTodos() {
const result = await db.allDocs({ include_docs: true });
result.rows.forEach(row => {
// タスクを取得して表示
console.log(row.doc.text);
});
}
ToDo の削除
async function deleteTodo(id: string) {
const doc = await db.get(id);
// タスクを削除
await db.remove(doc);
}
Running the To-Do App
Open the application in a web browser.
Enter a task and click the Add button—the task will be added to the list below.
Close the browser. (Normally, this would cause the entered tasks to be lost.)
Reopen the application in the browser. (The previously entered tasks remain displayed.)
Click on a task to delete it.
Thoughts on Running the App
One of the standout features of this To-Do App is its ability to function independently of the internet, managing data entirely within the browser. By storing data locally, users can continue using the application even while offline.
Key Takeaways:
Data persists even after a page reload!
The app works without a server!
Fast performance with instant data retrieval!
Applications that don’t rely on servers or the cloud might seem somewhat uncommon, but the Local-First approach proves to be highly valuable for offline functionality and data privacy. While this project was a simple implementation, it could be extended with cloud synchronization or mobile support for a more versatile experience.
Exploring the possibilities of Local-First applications has been insightful, and I look forward to leveraging this concept further.
Welcome to our engineer blog, where we share our ongoing learning experiences. This post is part of Unit 0, which focuses on web design. Today, I’d like to introduce an opportunity I had to work with something less commonly encountered: QR code generation.
Preparation
For this implementation, I used Java. Since my development environment is Gradle-based, adding the necessary libraries was as simple as including the following in the build.gradle file. (Isn’t it convenient how accessible things have become these days?)
The library used here is ZXing, an open-source Java library that enables the creation and reading of one-dimensional codes (such as barcodes) and two-dimensional codes (like QR codes).
Implementation
The key element for generating QR codes is the encode method of the QRCodeWriter class. As summarized in the comments, the following parameters can be specified:
The generated data is stored in a variable of type BitMatrix (bm). By specifying the output format in the writeToStream method, you can save the generated QR code in your desired format. In this case, we output the QR code in PNG format.
While the actual implementation involves handling API requests and returning the output result to the screen, I’ll omit those details here.
Summary
Nowadays, if you just want to generate a QR code once for testing purposes, there’s no need to write a program from scratch. You can find many online QR code generators with just a quick search. Some even let you customize the design or offer formats tailored to specific use cases, making them surprisingly fun to explore.
This post covered only the basics, but I hope it gave you an idea of how QR code generation works. If you found this interesting, I’d be delighted.
Stay tuned for the next engineer blog post!
P.S. The content of the QR code includes a closing message!
As a tech blog, we’d like to introduce the learning content we’ve been working on daily. We are unit 1, focusing on AI and Big Data. In our previous article, we discussed the possibilities of content creation, education and learning, and business solutions using the chatbot “ChatGPT-4o”.
This time, we’ll focus on how we can streamline our main business of system development by incorporating AI into system testing to ensure quality.
1. Selenium 4
Overview:
Selenium is a widely used open-source test automation tool, particularly suitable for cross-browser testing. The latest version, Selenium 4, has improved APIs for manipulating UI elements, allowing for more intuitive test script creation. When combined with AI plugins, it also enhances the detection of dynamic UI elements.
Usage rate: High
Price: Free (open-source)
Difficulty level: Intermediate
Official website: Selenium Official Website
2. Cypress
Overview:
Cypress is a JavaScript-based test automation tool for web applications, particularly popular among frontend developers. It allows for real-time verification of test results and has a simple setup, making it easy for beginners to get started. It’s ideal for end-to-end testing and makes it easy to test complex user flows.
Usage rate: High
Price: Free plan available, premium plans (paid)
Difficulty level: Beginner to intermediate
Official website: Cypress Official Website
3. Katalon Studio
Overview:
Katalon Studio is a tool that allows for test automation without writing code, supporting a wide range of test scenarios including web applications, mobile apps, and APIs. With its GUI-based operation, even those unfamiliar with programming can intuitively create tests. It provides enterprise-level features while offering a free plan, making it widely used by beginners to advanced users.
Usage rate: High
Price: Free plan available, enterprise plan (paid)
Difficulty level: Beginner to intermediate
Official website: Katalon Studio Official Website
4. Applitools
Overview:
Applitools is a tool specialized in visual testing, where AI automatically detects changes in design and UI. It’s particularly strong in visual regression testing and is ideal for testing UI and UX consistency. Its ability to detect subtle differences in design helps in quality assurance for frontend and design.
Usage rate: High
Price: Paid (free trial available)
Difficulty level: Beginner-friendly
Official website: Applitools Official Website
In this article, we introduced AI-powered test automation tools for system development. In our next article, we’d like to share our experiences and impressions from actually using these tools. We hope this helps you consider which tool might be best for your project.
