Maximize GenAI Potential with a Low-Code Framework

Get started with GenAI and low code

We all know GenAI is important, and we don’t want to be left behind, but not everyone is explicitly in the business of technology adaptation. It’s additionally hard to devote resources to new technology when examples of AI in everyday life are still crystalizing for most people.

That’s just one reason why your use cases are so critical.

AI is different than previous deterministic systems and technology. Discriminating AI can learn, predict, and act. Generative AI’s creativity combines limitless possibilities with unpredictable results. Agentic AI makes decisions and executes tasks autonomously but requires careful oversight and control. But remember to check your ambition. When creating your first AI use case, you might learn more by building less.

Agentic AI makes decisions and executes tasks autonomously, but requires careful oversight and control.

Create a doable, measurable project first. Don’t let the crisis of so much choice obscure your AI-driven business objectives. A proper use case grounds an enterprise and separates your AI aspirations into the must-haves and the nice-to-haves.

Consider the following when creating your first AI-driven use case:

• What problem are you solving with AI? Don’t forget to define a successful solution with measurable success metrics.
• What AI features, services and products can help you solve your problem? What do your current systems do? What will AI need to do?
• What will the impact of AI integration be? What are the challenges you will face getting there? How will success change your enterprise for the better?

Want to preview your ROI? Test your AI use case with a quick-fire MVP prototype before establishing a proof of concept. Use low code and prebuilt AI models to limit the impact on your resources. Keep the scope as small as possible. Some examples to start with might be:

• Improved data quality: Validation rules, workflows enforcing data consistency, and Mendix’s GenAI Product Information & Asset Management (PIAM) solution are just the start.
• Better risk controls: From audit trails to integrated compliance management, Mendix low-code introduces next-level governance opportunities throughout the SDLC.
• Better efficiency: Mendix has a host of opportunities to reduce cost and time-to-revenue, from scaling to open-source development.
• Business value discovery: Faster, easier access to GenAI foundational models through connectors and APIs helps securely introduce proprietary data with fewer resources.

Low-code chatbots that exponentially improve CX

Mendix low-code partner enterprise Vivix Vidros Planos is a Brazilian-based commercial and residential glass producer. Vivix is finishing work on a second plant in Brazil that will double its capacity for production from 900 tons of flat glass to more than 1,900 tons, making it the second-largest producer in the country.

In 2021, as part of a three-year roadmap to operational digitization, Vivix deployed 17 low-code applications using the Mendix platform. The low-code apps recaptured more than 6,000 hours of manual work in 2023 alone. Building on the success of their earlier projects, Vivix is releasing a GenAI-powered chatbot experience for both customer complaints and internal production issues.

The Vivix chatbot experience

The Vivix Virtual Engineer uses the Mendix AWS Bedrock connector to invoke the Claude 3 foundation model from Anthropic. Virtual Engineer integrates data from several data sources to provide helpful, accurate guidance and insight when it’s needed most. Early Vivix estimates show just how valuable the AI-driven chatbot experience can be with data-driven insights, predictive behavior, and streamlined workflows. Time-to-resolution for customer-facing issues using the Virtual Engineer chatbot is estimated at less than a day—down from more than five days. Internal production issues are resolved in one hour—a five-fold reduction in resolution time.

According to Aristóteles Terceiro Neto, Industrial Transformation Manager at Vivix, “We want our team to use generative AI to improve our reports and have a new perspective of our data. We are also deeply committed to expanding the use of Mendix Edge applications through Siemens Industrial Edge.”

Enterprises can leverage one of several foundation models with the Amazon Bedrock connector and the Mendix platform, including:

• Amazon Titan
• Anthropic Claude
• Cohere Command & Embed
• AI21 Labs Jurassic
• Meta Llama 2
• Mistral AI
• Stability AI Stable Diffusion XL

Enterprises that use the Mendix low-code platform can also integrate their internal data with foundation models from sources other than Amazon Bedrock with connectors available in the Mendix Marketplace. OpenAI’s ChatGPT, for example, is available through CapGemini’s ChatCPT connector.

Remove complexity with more personalized user experiences

Modern sustainability initiatives are founded upon cutting-edge technology. For the average homeowner, conducting the necessary research, knowing a fair price, and finding reputable vendors and sustainability service providers is a full-time job. With ambitious international environmental goals at both the commercial and residential building levels, connecting and completing residential upgrades is even harder as schedules fill up.

HomeServe’s marketplace model connects homeowners with contractors to complete work. In addition, HomeZero provides a platform that makes it easy for installers to build profiles and connect with potential customers, forging bonds from both angles. In this way, much of the technological complexity of innovation is diffused, replaced by the opportunity for a third-party connection.

GenAI makes it easier to make good choices

In addition to making the introduction between customers and potential contractors, HomeZero helps homeowners understand what solutions are best for their needs. With the help of GenAI, residents receive real-time advice and contractor suggestions from a virtual sustainability officer based on their prompt input and the information they provide about the project.

According to HomeZero founder Mike Out, “The beauty is that we have a managed solution via Mendix. We don’t see all the complexities of AWS, but we gain all the benefits…We have all these niche ideas, products, and prototypes that we’re building out now for our users. There are so many possibilities to transform this industry even more.”

Future-proof security protocols

ABN AMRO is the third-largest bank in the Netherlands. As a financial institution in the digital age, ABN AMRO understands the challenges of innovation in a space that requires stable, secure, and reliable software.

Guided by a three-pillar strategy of customer experience, sustainability, and future-proof banking, ABN AMRO used Mendix low code to rebuild legacy systems. With more than 150 low-code apps in production, ABN AMRO has plans to migrate hundreds of operations and services to their low-code platform by the end of the year. They are enacting their strategy through a framework of governed rapid application development and low-code.

Safer banking through GenAI

As part of the ABN AMRO financial crime detection suite of tools, GenAI is used to perform compliance measures, such as maintaining best practice Know Your Customer (KYC) standards. The process starts with access to relevant data sources from which useful customer profile summaries get created through GenAI. These profiles accelerate customer assessment and security protocol processes.

According to Benjamin Blaauw, the head of development automation at ABN AMRO, “Mendix is a target technology to simplify our landscape and speed up development. … If we see there is a short timeline and we need to deliver something quickly, then Mendix is the way forward.”