How low-code platforms enable machine learning

【51CTO.com Quick Translation】

[[425497]]

Low-code platforms improve the speed and quality of developing applications, integrations, and data visualizations. Instead of building forms and workflows in code, low-code platforms provide a drag-and-drop interface to design screens, workflows, and data visualizations used in web and mobile applications. Low-code integration tools support data integration, data preparation, API orchestration, and connections to common SaaS platforms. If you are designing dashboards and reports, there are many low-code options for connecting to data sources and creating data visualizations.

If it needs to be done with code, there may be a low-code or no-code technology that can help speed up the development process and simplify ongoing maintenance. Of course, the platform must be evaluated to see if it can meet functional requirements, cost, compliance, and other factors, but the options offered by low-code platforms fall in the gray area between building your own or purchasing a software as a service (SaaS) solution.

Are low-code options just for better and faster development of applications, integrations, and visualizations? What about low-code platforms that are accelerated and simplified with more advanced or emerging capabilities?

The following is a detailed explanation of how low-code and no-code platforms enable technical teams to leverage machine learning capabilities for testing.

Platforms for different development roles

For data scientists, is it exciting to use new machine learning algorithms that support model operations faster and easier than Python coding through low-code capabilities? Or for data engineers, focusing on data operations, looking to connect data to machine learning models while discovering and validating new data sources.

Data science and model operation platforms, such as Alteryx, Dataiku, DataRobot, H20.ai, KNIME, RapidMiner, SageMaker, SAS, etc., are designed to simplify and accelerate the work of data scientists and other data professionals. They have comprehensive machine learning capabilities, but are easier to use for professionals with data science and data engineering skills.

Here’s what Dr. Rosaria Silipo, Chief Data Scientist and Evangelist at KNIME, said about low-code machine learning and AI platforms. “AI low-code platforms are a valid alternative to traditional AI scripting platforms. By removing the coding barrier, low-code solutions reduce the learning time required for tools and leave more time for experimenting with new ideas, paradigms, strategies, optimizations, and data.”

There are multiple platform options, particularly for software developers looking to leverage machine learning capabilities in their applications and integrations:

• Public cloud tools, such as GCP AutoML and Azure Machine Learning Designer, help developers access machine learning capabilities.
• Low-code development platforms, such as Google’s AppSheet, Microsoft’s Power Automation’s AI Builder, and OutSystems’ ML Builder, expose machine learning capabilities.
• Low-code learning libraries such as PyCaret target data scientists, citizen data scientists, and developers to help accelerate learning and implementation of machine learning on open source toolkits.

These low-code examples target developers and data scientists with coding skills, helping them accelerate experimentation with different machine learning algorithms. The MLops platform targets developers, data scientists, and operations engineers. The MLops platform effectively supports devops for machine learning and is designed to simplify managing machine learning model infrastructure, deployment, and ops management.

No-Code Machine Learning for Analysts

An emerging group of no-code machine learning platforms are targeted at business analysts. These platforms make it easy to upload or connect to cloud data sources and experiment with machine learning algorithms.

I spoke with Assaf Egozi, co-founder and CEO of Noogata, about why a no-code machine learning platform for business analysts is a game changer even for large enterprises with experienced data science teams. He told me, "Most data consumers within an organization simply don’t have the skills required to develop algorithms from scratch or even effectively apply autoML tools — and we shouldn’t expect them to do so. Instead, we should provide these data consumers — citizen data analysts — with an easy way to integrate advanced analytics into their business processes."

Andrew Clark, CTO and co-founder of Monitaur, agrees. “It’s exciting to make machine learning more accessible to the enterprise. There aren’t enough trained data scientists or engineers with the expertise in productizing models to meet business needs. Low-code platforms provide a bridge.”

While low-code democratizes and accelerates machine learning experimentation, it still requires rigorous practice, alignment with data governance policies, and review for bias. Clark added, “Companies must view low-code as a tool to benefit from AI/ML. They should not take shortcuts given the business visibility, control, and model management required to make trusted decisions for the business.”

Low-code capabilities for software developers

Now let’s focus on low-code platforms that provide machine learning capabilities to software developers. These platforms select machine learning algorithms based on their programming model and the types of low-code capabilities they expose.

• Appian provides integration with several Google APIs, including GCP Native Language, GCP Translation, GCP Vision, and Azure Language Understanding (LUIS).
• Creatio is a low-code platform for process management and customer relationship management (CRM) with a variety of machine learning capabilities, including email text mining and universal scoring models for leads, opportunities, and accounts.
• Google AppSheet supports a variety of text processing functions, including intelligent search, content classification, and sentiment analysis, while also providing trend prediction. After integrating data sources such as Google Sheets, you can start experimenting with different models.
• The Mendix Marketplace has machine learning connectors to the Azure Face API and Amazon Rekognition.
• Microsoft Power Automate AI Builder has features related to processing unstructured data, such as reading business cards and processing invoices and receipts. They use several algorithms, including key phase extraction, category classification, and entity extraction.
• OutSystems ML Builder has a variety of capabilities that may come up when developing end-user applications, such as text classification, attribute prediction, anomaly detection, and image classification.
• Thinkwise AutoML is designed for classification and regression machine learning problems and can be used in a scheduled process.
• Vantiq is a low-code, event-driven architecture platform that can power real-time machine learning applications such as AI monitoring of factory workers and real-time translation of human-machine interfaces.

This is not a comprehensive list. A list of low-code and no-code machine learning platforms also named Create ML, MakeML, MonkeyLearn Studio, Clearly AI, Teachable Machine, and other options. Also, take a look at the no-code machine learning platforms and no-code machine learning platforms for 2021. This is increasingly likely as more low-code platforms develop or partner to develop machine learning capabilities.

When to use machine learning capabilities in low-code platforms

Low-code platforms will continue to differentiate their feature sets, so I expect more platforms will add the machine learning capabilities needed for the user experiences they enable. This means more text and image processing to power workflows, trend analysis for portfolio management platforms, and clustering for CRM and marketing workflows.

But when it comes to supervised and unsupervised learning, deep learning, and model operations at scale, it is more likely that specialized data science and model operations platforms will need to be used and integrated. More low-code technology vendors may collaborate to support integrations or provide on-ramps to enable machine learning capabilities on AWS, Azure, GCP, and other public clouds.

What continues to matter is that low-code technologies make it easier for developers to create and support applications, integrations, and visualizations. Now, raise the bar and expect more intelligent automation and machine learning capabilities, whether low-code platforms invest in their own AI capabilities or offer integrations with third-party data science platforms.

[Translated by 51CTO. Please indicate the original translator and source as 51CTO.com when reprinting on partner sites]