Jan Arendtsz, Founder and CEO of Celigo – Interview Series

Jan Arendtsz is the Founder and CEO of Celigo and a veteran of the software industry with more than 25 years of experience in product development, business development, sales, customer success, and marketing. He founded Celigo with the goal of simplifying how companies integrate, automate, and optimize business processes across the enterprise. He is responsible for overseeing all company operations.

Prior to Celigo, Jan was a Director at NetSuite, the leading cloud-based ERP platform, where he launched their integration platform. Prior to that, Jan worked for Cambridge Technology Partners, where he implemented complex business solutions for customers ranging from internet startups to Fortune 500 companies.

Celigo is a cloud-based intelligent automation and integration platform (iPaaS) designed to help organizations connect applications, automate business processes, and keep data synchronized across their technology stack without heavy custom development. Its platform combines prebuilt connectors, reusable integration templates, and AI-assisted tools so both technical and non-technical teams can design, deploy, and manage integrations at scale. Celigo is commonly used to streamline workflows across areas like ecommerce, finance, operations, and IT, reducing manual work, improving data accuracy, and enabling faster, more resilient business processes.

What originally motivated you to found Celigo, and how did your experience leading integration services and product initiatives at software companies like NetSuite shape the gap you saw in enterprise integration at the time?

That early SaaS experience showed me that while the cloud solved a software delivery problem, it also created a massive data connectivity problem. We were selling the vision of a unified business, but the reality was fragmented data silos. I started Celigo to fix those challenges.

Today, I see history repeating itself with AI. We are shifting from a “connectivity gap” to an “operational gap.” Just as companies struggled to operationalize SaaS twenty years ago, they are now struggling to operationalize AI. Companies are struggling with moving AI from experiments to reliable business outcomes. This creates the next wave of challenges that Celigo is uniquely positioned to help IT leaders address: how to provide a platform that doesn’t just connect systems but enables the use of AI across their enterprises at scale.

Celigo has evolved from traditional integration toward AI-driven workflows. What signals told you the platform needed to move in that direction?

The biggest signal was the shifting bottleneck. Ten years ago, the bottleneck was connectivity: just getting System A to talk to System B. We solved that with iPaaS. But as we democratized integration and empowered business users to build their own workflows, the new bottleneck became management, governance, and exception handling.

We looked at our data and saw that while building automation workflows had become easier, maintaining them at scale was still human-intensive. Users were spending hours troubleshooting data errors or updating mappings.

We responded by embedding AI into the core of our platform, automating error classification and remediation to remove the operational burden of maintaining integrations at scale. That platform intelligence now sets the stage for customer-facing AI-powered workflows that can operate with greater autonomy and context.

Many organizations are investing heavily in AI but seeing limited results. Why do so many initiatives stall at the data and integration layer?

We’ve all seen the surveys showing that while most companies are experimenting with AI, few are realizing measurable ROI. The reason isn’t the technology. It’s the approach. Too often, organizations treat AI adoption as the objective rather than starting with the business processes that run the enterprise.

Successful initiatives begin by identifying the processes where improvement drives the biggest business impact, rather than applying AI to isolated tasks. From there, AI must be connected to the systems where work actually happens, with guardrails that ensure data quality and policy enforcement. Without this governed connectivity, AI remains detached from execution.

Finally, AI requires an orchestrated framework that balances autonomy with control. Human-in-the-loop workflows and exception handling are critical for maintaining trust as AI takes on more responsibility. When AI is embedded into end-to-end business processes, it evolves from a novelty into an operational enabler that delivers real business results.

From your perspective, what architectural mistakes are most common when companies try to layer AI on top of fragmented systems?

A growing problem right now is AI sprawl. We often see companies purchasing multiple different SaaS extensions: sales tools with AI, customer service tools with AI, marketing tools with AI, and so on. These are all just wrappers around the same underlying LLMs.

Architecturally, this can create significant cost and governance issues. IT leaders are finding that they need an integration platform to be able to consolidate all of the data and insights across tools. Leveraging a unified platform can bring together the knowledge that resides across the enterprise and provide the context needed for AI models to scale and deliver value.

As AI becomes more autonomous, how do intelligent workflows change the way applications, data, and people interact inside an organization?

As AI becomes more autonomous, intelligent workflows change how applications, data, and people interact by shifting automation from task execution to decision orchestration. Applications are no longer just connected to exchange data. They become coordinated participants in a workflow where AI interprets context across systems and determines the next best action.

This shift brings change management to the forefront. You can have the best model in the world, but if teams don’t trust it, they won’t use it. Successful operationalization of AI requires visibility into why an AI agent made a specific decision and confidence that it is operating within a governed framework.

As workflows evolve from performing tasks to reviewing outcomes, people move from being operators to overseers. Users can choose the level of autonomy they entrust to AI, with human in the loop controls providing governance, accountability, and adaptability as agents improve over time. The result is a dynamic, hybrid environment where applications act, AI decides, and people guide.

Celigo serves both large enterprises and fast-growing brands. How do integration, data quality, and orchestration challenges differ across those stages of scale?

For fast-growing brands, the goal is often speed to value. They are adopting tools so fast that they risk building a fractured stack that will break in a year or two. For them, Celigo offers the ability to operationalize quickly without creating technical debt.

For large enterprises, the challenges are around context and governance. They have valuable stores of data, but it may not be ready for AI-led workflows. They need to increase the access and value of data across complex environments. They need to ensure that as they operationalize AI across their organizations, they aren’t leaking PII or hallucinations into their customer interactions. We serve as an important management and control layer.

Celigo sits at the intersection of iPaaS, workflow orchestration, and AI. How should organizations design their integration layer so it becomes an active part of the AI stack rather than passive infrastructure?
Organizations are moving beyond integration as simple data movement toward intelligent automation as a controlled connectivity layer for business processes. Automation exists on a spectrum from predictable, rules-based execution to more autonomous behavior, with the greatest enterprise value created in the middle.

An intelligent automation platform connects AI to the right enterprise data with built-in governance, visibility, and human oversight. It orchestrates connectivity across systems, applies intelligence selectively, and executes outcomes directly within operational applications where work actually happens. Rather than passively moving data between systems, the integration layer becomes active by maintaining real-time, governed connectivity and control. This ensures intelligent automation remains reliable, auditable, and aligned with how the business is designed to run.

With the rise of agentic AI, what role do you see integration platforms playing in enabling AI systems to take action safely and reliably across business applications?

Agentic AI needs guardrails. Celigo is building a future where integrations will increasingly manage themselves by detecting schema changes, predicting failures, and self-healing before a human even knows something is wrong.

The role of our platform is not only to allow users across the business to build and run workflows quickly and efficiently, but also to allow central IT to provide those guardrails. If an agent wants to update a record, the platform will ensure that action validates against business rules first. We enable agents to take action by providing a deterministic environment where non-deterministic AI can operate safely.

Looking ahead to 2026, what do you see as the real-world consequences for organizations that fail to streamline their data specifically for AI?

The consequence will be a divergence in ROI. Companies that fail to embed AI into their operations will be limited by measuring “hours saved” on ad-hoc tasks, while their competitors will be measuring “revenue growth” from fully automated lines of business.

Without taking the steps now to connect their data and applications, organizations will hit a wall where AI models are hallucinating because they lack context, or costs spiral because they’re paying for the siloed outputs of AI sprawl without a unified strategy. Companies could effectively be priced out of agility.

For technology leaders modernizing their stacks today, what core capabilities should they prioritize in an iPaaS to ensure their AI initiatives can scale and deliver real outcomes?

Look for a modern iPaaS that was built for a world where everything needs to connect with everything. That means: a universal platform that can handle the full spectrum of automation: from data and application integration, to B2B supply chain flows, API management, and autonomous agents. That sets an organization up for less complexity, less overhead, more empowerment of users, and ultimately the ability for IT to strategically and safely incorporate AI and operationalize everything across the enterprise.

Thank you for the great interview, readers who wish to learn more should visit Celigo.