Which Low-code development platform is best?

Which Low-code development platform is best?

Are you grappling with deciding which low-code application development platform to use? The landscape of low-code (or no-code) development, which has been around for decades, has recently exploded with many options. However, selecting a low-code platform is not a straightforward decision. Here's what you need to carefully consider before making this significant investment.

Broadly seen, there are three major types of low-code platforms:

  1. The "Built-in" low-code toolsets for automation
  2. Add-on integration, workflow, and automation platforms
  3. Low-code programming platforms for building complex systems and enterprise-scale applications

?If you know much about these low-code platforms, please skip to My take below.

?Types of Low-code platforms

Built-in low-code toolsets are the most common, where the tools are built right into a specific software package or service. Almost all major enterprise software platforms - Salesforce, ServiceNow, SAP, Atlassian - can build process flows and automation within their confines. These toolsets are captive in that they are only meant to extend and customize the central platform rather than provide a way of building automation or even applications from scratch.

Add-on low-code tools provide the means of building workflow applications that integrate data and services from other applications. They often run standalone but have a rich set of connectors—pre-built integrations with different platforms and applications. These low-code platforms can vary widely in size and scope and often include additional features such as Robotic Process Automation. On the larger side, IPaaS platforms like Pega and Workato allow organizations to build large-scale integration flows using drag-and-drop development tools.

One Add-on automation toolset that has Built-in characteristics is Microsoft's Power Apps platform, with Power Automate at its core. As Power Apps is native to the Microsoft 365 (ne Office 365) platform, it focuses on automation and front-end UIs that extend and integrate the various Microsoft cloud properties.?

For building large, complex enterprise applications with multiple UI channels that need to scale up, full-stack low-code development platforms like OutSystems and Mendix provide a visual programming approach incorporating tools to accelerate and simplify the application-building process. They are "full-stack" platforms that support all layers of modern applications, from data and integration to expressive and responsive user interfaces.

Which low-code type should you use?

Of course, as a consultant, I will always say, "It depends." Nevertheless, here's some guidance on deciding what to adopt.

Are you heavily focused and invested in a single software platform like Salesforce, Workday, or ServiceNow? If so, their built-in low-code platform(s) is sufficient for your needs, including integrating your core platform with other systems.

Do you have a large heterogeneous set of applications and services you'd like to integrate? Will automating processes involve multiple disparate systems? If so, then an Add-on low-code platform may be a good fit for your needs.?

Are you looking for ways to automate the processing of existing assets in your Microsoft 365 environment, such as those in SharePoint or Dynamics? Do you need to build simple UIs that provide better access to data across your Microsoft and related systems? Power Automate and Power Apps would then be a natural starting point.

Do you want to build applications that integrate existing applications and provide a better way of accessing and using them? Do you need to modernize your application portfolio? Are there custom applications you want to build that provide unique competitive advantages rapidly? Adopting a full-stack low-code platform like OutSystems can be an excellent short-term and long-term choice.

For larger organizations, it's important to note that you're not limited to just one of these low-code platforms. In fact, having a mix of these platforms can be beneficial, as each one is well-suited for its specific use case. This flexibility empowers you to choose the best tool for each job.

My take

Power Apps/Automate is a fine starting point, but it won't scale

If all you need is basic back-office automation, if you have Microsoft 365, Power Apps is a great place to start building automation. It's straightforward for simple to moderate automation, easily integrates with your Microsoft-based assets, and has reasonable costs.

However, I don't recommend using Power Apps for new, complex applications or integrations. The visual programming language is not designed for building more extensive applications, as building Power Automate flows longer than a few screens becomes extremely hard to read and maintain. Debugging large flows becomes tenuous, and true modularity is impossible as flows cannot call other flows.

Don't expect to build large applications using your existing enterprise applications.

Using large, typical enterprise applications like SAP, ServiceNow, and Salesforce makes much sense for organizations needing those capabilities. However, none of those applications can provide any strategic advantages to your business. They can offer no unique 'special digital sauce' that will differentiate how you do business from your competitors. The low-code capability of these platforms is not meant for new applications; they only extend what's already there.

To build well-tailored applications and incorporate your special sauce, consider using a Low-code programming platform like OutSystems. These platforms provide a full-stack solution that allows you to create modern web and mobile applications and quickly integrate other data and web services. With the right team and guidance, you can build a production system with unique capabilities that outshine the competition and enable rapid business growth.

Why not stick with high-code?

Many traditional developers will tell you they can code applications as fast using traditional 'high code' (text-based programming languages) as any low-code platform. Highly experienced programmers may perform extremely fast, but building more complex applications usually requires teams of developers with an end-to-end DevOps process in place first. Platforms like OutSystems come with DevOps built-in, along with a pre-structured application architecture that is proven to scale and be secure.

A few years ago, I was a traditional developer. But after seeing my team rapidly develop—and extend—a system that has allowed a new business to grow rapidly, I've changed my tune. The value of low-code platforms is real.

No-code is a myth.

There are "No-code" platforms where apps and automation supposedly can be built without needing to learn a single programming language. Some No-code vendors imply that no programming or development skills are required to build apps. This is only true for platforms that heavily restrict what can be built by the user, focused on a handful of specific use cases.?

Building business applications with a rich set of functionality means implementing logic, process flow, and data models, regardless of what the 'coding' looks like. To state the obvious, developing applications requires understanding Application Development.

The additional reality is that any reasonably complex enterprise application may require extending the development platform. Low-code platforms can incorporate "high-code" elements for specific needs. For example, to better support real-time audio AI transcription in a medical application built with OutSystems, we embedded Javascript code in the reactive web page to take advantage of the browser's WebRTC capabilities.

In my experience, enterprise applications built on OutSystems typically involve 95-98% low-code visual programming, with a tiny percentage of elements built in Javascript, C#, or custom SQL, usually in support of integration with other systems.

The Reality of Low-code

We have realized excellent benefits from adopting Low-code, using OutSystems as the core application platform. Low-code has enabled us to increase our speed to production while maintaining a high level of quality. It has helped us develop applications with a team of four developers that would have taken a dozen to write traditional code.

However, Low-code platforms face a stiff challenge: getting companies to invest in building great applications. Most business executives want 'solutions'—pre-built software packages (or services) that provide them with all the functionality they need to run a business. Those pre-built solutions can be helpful but do nothing to give a company an edge over its competition. Building bespoke, best-fit applications that go beyond packaged functionality is the best way for companies to jump ahead of the pack.

As with any architectural decision, using low-code to build applications does have trade-offs, such as the lack of certain object-oriented language features. However, we have found that those trade-offs are worth the benefits. With OutSystems, many of the free components in their Forge library help address those trade-offs.

Low-code platforms operate like SaaS applications, where all the infrastructure costs are rolled into subscription costs. Compared to open-source programming tools, low-code can appear expensive. But building applications in high-code, even open source, requires significant effort in building, maintaining, and paying for DevOps, security, runtime infrastructure, database platforms, and many other components needed to care for and feed those applications. The resulting costs are often hidden within the extra hours, piecemeal payments made for individual tools, and the infrastructure costs in other budgets. Using a low-code platform lets your dev team focus on delivering high-value applications quickly without having to deal with having to build out (and pay for) the underlying support mechanisms.

Choosing to go Low-code

Do you need to find a better way to build applications? Would an integrated UI across multiple apps improve your business capabilities? Are you drowning in lines of code? Does your business have a 'secret sauce' that can't be bought in a software package? All these questions indicate that you would benefit from adopting a low-code platform.

But be realistic about Low-code. The benefits don't show up instantly. Getting the most out of a low-code platform requires learning how to use it effectively and applying good architecture and development practices. Low-code or not, solid software practices and principles are always essential. Low-code just makes applying them easier.

That said, I'm still going to program some things in C#, Python, C++, Go, JS/TS, and, if I get the chance again, LISP. But that's a story for another time. But I'll admit that I'm now biased towards OutSystems - not just because T4S Partners offers professional services with OutSystems, but also because it's a lot faster to get a great application out the door.

#LowCode #DigitalTransformation #OutSystems #Innovation #TechSolutions #BusinessGrowth

Felipe Andrés Barrera Lobo

Estrategia, Crecimiento y Nuevos Negocios | Startups | Aceleración y gestión de innovación corporativa y emprendedora | Levantamiento de capital público y privado

1 个月

One often overlooked factor in low-code platform selection is AI integration capabilities. I've seen platforms that handle AI components smoothly, while others need complex workarounds that defeat the whole "low-code" purpose.

要查看或添加评论,请登录

Rick Munoz的更多文章

  • Why We Need LLM Output Gatekeeping - Policing AI Agents

    Why We Need LLM Output Gatekeeping - Policing AI Agents

    In Computerphile's recent video on "Generative AI's Greatest Flaw," Mike Pound shows how indirect prompt injection…

  • Dividing Event-Sourcing into nanoservices

    Dividing Event-Sourcing into nanoservices

    A few years ago, I led the design of a cloud-native financial management system meant to run globally, with users in…

  • DeepSeek's AI improves when rewarded

    DeepSeek's AI improves when rewarded

    Reinforcement Learning (RL) has been a popular approach to training and improving AI models. Here's an…

    2 条评论
  • AI can't learn without asking

    AI can't learn without asking

    I ran across this great video from IBM Technology on "Can AI Think? " It points out that LLMs (and neural networks in…

    1 条评论
  • Meta AI's LCM: a good evolutionary step

    Meta AI's LCM: a good evolutionary step

    Meta's recent paper on Large Concept Models (LCMs) is being heralded as the next generation of AI technology. This is…

  • Three Keys to Successful App Transformation

    Three Keys to Successful App Transformation

    How old are your applications? Your database systems? Is any of your technology nearing (or way past) its end-of-life…

社区洞察

其他会员也浏览了