Vivix, a leading glass manufacturer, is transforming how its operations run by moving beyond traditional automation toward a more connected and scalable way of working.

What started in 2021 as a push to modernize applications with low code has evolved into something much broader. Rather than simply digitizing processes, Vivix rethought how information flows across the organization and how teams interact with it day to day.

Along the way, Vivix has been recognized for its innovation with the 2024 Techcellence Award for Innovation, the AWS Gen AI Gamechanger award, and the 2026 Techcellence Award for Digital Transformation.

At Realize Live 2026, Aristóteles Terceiro Neto, Industrial Transformation Manager at Vivix, shared how the team is scaling its Mendix portfolio, now approaching 30 applications, while building the foundation needed to support AI at scale.

The agentic enterprise isn’t a concept at Vivix–it’s how the business operates.

Enhancing the Industrial Data Fabric

For Vivix, the biggest challenge wasn’t a lack of technology but building on earlier digital and data initiatives to make complex industrial data truly usable at scale.

Like many manufacturers, they had data everywhere: on the shop floor, at the edge, and across enterprise systems. But connecting that data in a meaningful way and turning it into action remained a challenge.

“Data contextualization is the hardest part. Start with the business problem, not with AI tools, and focus on a reusable architecture,” Neto emphasized.

Instead of jumping straight into AI, Vivix focused on building the right foundation first.

They developed an industrial data fabric: a unified architecture connecting systems across the operation, including HighByte, Snowflake, AWS, Siemens technologies, Altair, and Mendix as the application layer.

This foundation enables Vivix to:

• Connect shop floor, edge, and cloud systems
• Standardize how data is modeled and accessed
• Ensure data quality, lineage, and governance
• Reuse components across use cases

“We use a unified structure that shares data, analytics, and security to integrate everything,” he added.

From Applications to Measurable Impact

With the foundation in place, Vivix quickly shifted to delivering applications that solve real operational challenges and drive measurable results.

“The results are very clear and tangible,” shared Murilo Ferreira, Senior Industrial Transformation and Digital Squad Leader at Vivix.

Using Mendix, Vivix has delivered more than 20 enterprise applications across maintenance, logistics, and production. These applications are embedded directly into daily workflows and tightly integrated with operational data.

One of the first areas to benefit was maintenance. Previously, teams relied on fragmented systems and manual analysis to diagnose issues, often slowing down response times and increasing risk. With Mendix, Vivix built applications that bring together operational data, historical context, and decision support into a single environment.

Instead of spending hours, or even days, understanding what went wrong, teams can now identify root causes in a fraction of the time. This shift has led to an 85% reduction in root cause analysis time, while also saving more than 2,000 hours annually in planning and troubleshooting. By anticipating failures before they happen, Vivix is also driving over $1 million in annual savings through predictive maintenance.

A similar transformation is happening in logistics. Managing reverse supply chain operations, like tracking and returning glass racks, was previously a complex and time-consuming process with limited visibility. By embedding AI into logistics workflows, Vivix has introduced real-time insights and recommendations that help teams make faster, more informed decisions.

The impact has been immediate: freight contract costs have been reduced by 50%, while turnaround time for glass rack returns has improved by 10 days. What was once a reactive process is now proactive and continuously optimized.

In production, applications built on Mendix give operators and engineers real-time visibility into operations, enabling them to respond quickly to anomalies and maintain process stability. Rather than relying on delayed reporting or manual checks, teams now operate with up-to-date insights directly embedded in their workflows resulting in faster decisions, fewer disruptions, and more consistent output.

Across each of these areas, the pattern is the same: applications are not just exposing data, they are actively shaping how work gets done.

Embedding AI into the Flow of Work

As these applications scaled across the business, Vivix began taking the next step of embedding AI directly into the workflows they had already transformed.

Instead of treating AI as a separate capability, Vivix integrates it into the processes people already use, making it a natural part of how work gets done.

“These agents are embedded into our operational workflows, not just personal productivity tools. Sometimes they automate, sometimes they help in decisions and increase productivity,” Neto said.

These agents:

• Automate repetitive operational tasks
• Provide real-time recommendations
• Support decision-making across maintenance, logistics, and production
• Identify patterns and risks early

Because they operate inside Mendix applications, they are directly tied to business processes, not disconnected tools. This approach has enabled rapid scale: “We implemented more than 22 enterprise apps and 85 fully digitalized processes,” Neto said.

Driving Innovation with AI and Digital Twins

Building on this foundation, Vivix is also applying AI to more advanced use cases. One of the most impactful examples is their digital twin initiative, which models critical industrial processes such as furnace operations.

“We created a digital twin of our furnace which increased energy efficiency by 5%. That translates to millions in savings and extends a furnace life from 12–14 years to up to 20 years,” Neto highlighted.

By combining machine learning with physics-based modeling, Vivix can:

• Simulate operating conditions
• Predict performance outcomes
• Optimize energy consumption
• Extend asset lifecycles

This goes beyond traditional monitoring. It enables a deeper level of operational intelligence that directly impacts cost, efficiency, and sustainability.

Scaling AI with Governance Built In

As the number of applications and agents grows, governance becomes critical. For Vivix, this was never an afterthought.

“Scaling requires governance, architecture, and reusable foundation from the beginning. Our main challenge is creating agentic solutions that are already governed,” Neto said.

Using Mendix, Vivix created a centralized environment where they can:

• Monitor and manage applications and agents
• Track performance and outcomes
• Ensure data consistency and quality
• Control how and where AI is applied

At the same time, human oversight remains essential. “We needed to identify when to keep humans in the loop, who allows an action to take place,” he added.

In many industrial scenarios, decisions have real-world consequences. By embedding human-in-the-loop controls into workflows, Vivix ensures that AI enhances decision-making without removing accountability.

A Flexible Platform Strategy

While Mendix serves as the core platform, Vivix takes a flexible approach to AI and system architecture.

“The decision depends on the complexity and the number of agent interactions,” Neto explained. “For simple architectures, normally we use Mendix. When it becomes very complex, like many agents interacting, we use Amazon. But Mendix is where most of our applications live.”

This approach allows Vivix to move quickly while still handling complexity, using Mendix for rapid development and orchestration, and bringing in advanced AI capabilities when needed.

The result is a platform strategy that balances speed, flexibility, and scalability.

Growing the Agentic Enterprise

Vivix is showing what becomes possible when organizations move beyond isolated use cases and invest in a shared, scalable foundation for AI.

With Mendix, they’ve built a platform that connects data, applications, and agents that are enabling continuous improvement across operations.

From predictive maintenance to digital twins and logistics optimization, their approach delivers measurable results while setting the stage for future growth.

“AI helps our team quickly find information, choose the best strategy, and take action much faster than before,” Ferreira added.

As Vivix continues scaling toward hundreds of embedded agents, their model is becoming clear: intelligence is no longer separate from operations—it’s built into them.