Q&A with Zinier CPO Andrew Wolf: Agentic AI and the Future of Field Service

I recently sat down with Zinier's Chief Product Officer (CPO), Andrew Wolf, to discuss Zinier's focus on Agentic AI and how it will revolutionize the field service management (FSM) industry.

If you missed the whitepaper, written by Andrew himself, you can download it here: Field Service AI Agents

Here’s what he had to share:

What made you realize AI agents were the right approach for field service challenges?

Andrew: The biggest thing for me was the automation piece. We’ve always been thinking about automation since Zinier’s inception. The core idea has always been: what parts of field service could be automated? If you can automate the routine, the only thing people are left doing is managing by exception. That frees you up to focus on improving the customer experience, achieving truly touchless service.

When OpenAI launched ChatGPT in November 2023, we did a prototype and tested it right away. But GPT and LLMs on their own aren't fully automated; they're not totally touchless. The real turning point, the moment we thought, "Oh wow, something interesting can happen here," was when we saw GPT could actually trigger APIs using LLMs. It could take action.

It's just a beautiful marriage with our event-based architecture and workflows. We realized this wasn't just another feature; it was an entirely new approach to service delivery that aligns perfectly with Zinier’s architecture and capabilities.

"If you can automate the routine, the only thing people are left doing is managing by exception. That frees you up to focus on improving the customer experience, achieving truly touchless service."

In your whitepaper, you talk about "touchless" service. What's the biggest challenge in making that a reality?

Andrew: When it comes to touchless field service, the first hurdle is getting the entire operational workflow digitized and set up within the system. You can’t just jump from zero to touchless overnight. Digitization is step one; touchless automation is the second step we're trying to layer on top.

Once you digitize, once you get the workflow in place, then you can start thinking about where things can become automated and touchless. Scheduling is a good example. You start by layering on recommendations, then move towards handling things purely by exception. Our Recommendation Center is a fantastic tool to help companies move towards touchless operations, and by extension, so are AI agents.

So, the challenge isn't just the technology; it's understanding the solution during discovery, identifying those opportunities to automate, and then managing the organizational shift. There's a lot of muscle memory built around human-centered service. It takes time to start thinking in this new way. But those willing to undertake it will see huge step-changes in productivity, not just incremental improvements. It's about redefining roles, which is a significant paradigm shift.

What makes Zinier's platform especially good for supporting AI agents?

Andrew: Two main things. First, we really did build an event-based, workflow-driven architecture from the ground up. Agents often operate in the background, listening for some trigger or event. Our entire architecture supports that inherently. It allows us to easily connect to external models or externally built agents.

Second is our platform's extensibility – the ability to modify the system, integrate easily, and leverage that workflow and event-driven architecture. This emphasis on automation is ingrained in our DNA. Our desire has always been to improve the lives of deskless workers. What can we automate for them? How can agents help field workers spend less time navigating systems and more time solving customer problems? Improving deskless workers' lives is fundamental to us, and agents are the next tool to help us take that to the next level.

Other solutions might be trying to shoehorn AI into archaic systems. Because our platform was based on workflows from the start – essentially flowcharts where actions happen based on events occurring, and events can trigger multiple workflows – we can incorporate LLMs and AI agents easily and move really quickly.

Can you share a simple example of how Zinier's AI agents might work together to solve a real problem?

Andrew: Absolutely. Imagine a scenario – maybe based on a real customer situation – where a specific piece of calibration equipment requires, say, three bearing replacements in a short period.

An AI agent monitoring contracts could flag that the customer is approaching their spend limit for that asset type and suggest an update or review. Simultaneously, a parts agent, noticing the increased consumption, could proactively increase the stock level for that specific bearing for that customer's typical service locations. And critically, a workflow agent could analyze the repeated failures and automatically add a new preventative maintenance check for bearing wear to the standard maintenance protocol for that equipment model.

The key here is that these agents work together, identifying a complex issue and proposing coordinated solutions across different operational areas – contracts, inventory, and maintenance procedures. All that might be needed is human approval, not deep intervention or analysis. This is the kind of intelligent, proactive service improvement that goes way beyond traditional FSM.

"Those willing to undertake it will see huge step-changes in productivity, not just incremental improvements. It's about redefining roles, which is a significant paradigm shift."

How do you see AI agents and human technicians working together in the future?

Andrew: We’re designing a future where AI agents handle the cognitive overhead that distracts technicians – and back-office workers, for that matter – from their most valuable work. Think about how much energy technicians currently spend just navigating different systems, searching for the right information or manual, documenting their work, managing logistics like parts ordering... all of that takes focus away from the core task: diagnosing and resolving the customer's technical issue.

Our vision is for AI agents to become the ultimate partner, almost an invisible one. They handle the background tasks, surface the right information at the right time, automate the documentation, and manage the logistics. It's absolutely not about replacing technicians. It's about removing everything that gets in the way of them doing their best work, making them more effective and focused on the high-value problem-solving they excel at.