Lessons From a Mid-Flight AI Transformation

I suddenly had more time for higher-value, strategic work. I was no longer bogged down in the invisible, repetitive tasks we all have. And I wasn’t just getting more done, I was thinking differently about my work. I felt like a stronger professional with a renewed sense of motivation and, to my surprise, a feeling of excitement about working with AI. It was helping me learn new skills and explore ideas I might never have made time for otherwise.

That was my personal “aha” moment.

Orium has been exploring automation and AI for a while—we even have patents to show for it. But over the past year, things shifted from interesting to urgent. That change motivated me, and many others, to move beyond the talk and start experimenting in real ways. And once we did, the energy started to spread.

But what I experienced was just one version of what was unfolding across the company. People meet change at different speeds, and Orians are no different. We knew we needed to help employees navigate their own AI journey, and we needed to do it fast.

That kind of exploration was more challenging in a tough economic climate, though. In the past, we could give people time and space to adapt, but with market uncertainty and rapid shifts in AI, we no longer had that luxury. We had to move quickly and find a way to support this major tech change while staying focused on the day-to-day work.

So we turned the AI imperative into an opportunity to reignite our spark, build momentum, and rally people together around something exciting.

As with any change, there was some initial resistance from both leaders and team members. But once a group of early adopters jumped in and began testing tools and sharing wins, their energy created momentum and helped bring others along. And from that wave of experimentation, the real lessons began to emerge.

This journey hasn’t been perfect. We’ve navigated transformations before, but AI presents a different kind of challenge, and we’re learning, adjusting, and refining our approach as we go. What’s kept us moving is our commitment to the team: listening, adapting, and staying grounded in what makes Orium special— our people.

Here are some of the biggest things we’ve learned about embedding AI into daily work:

(1) Clarity is Table Stakes → Everyone needs to know the “why” and “how.” From the outset, we worked to provide clarity around our AI strategy. We created a manager hub with plain-language resources and clear steps leaders could take with their teams, ensuring consistent messaging across the organization.

But while that was helpful, we heard loud and clear that Orians wanted more guidance on the “why” and “how” of AI in their day-to-day operations. That tracks. If people don’t see what’s in it for them, adoption stalls.

To address this, we worked with functional leaders to define how their teams should collaborate with AI today, what experimentation could look like tomorrow, and how their roles might evolve as AI matures. By focusing on both the big picture and the day-to-day, we were able to reduce friction and keep adoption purposeful and actionable for our team members.

Of course, this is only the beginning. As AI becomes more embedded in our work, we’ll need to get increasingly specific about what it means for each role. AI will inevitably reshape and reconfigure how we work, and in some cases, displace parts of jobs altogether. But it will also create new roles and new opportunities. Our job is to help team members reimagine what their future at Orium could look like and equip them with the tools, skills, and confidence to get there.

(2) Match enablement to readiness → Right training, right time. Not everyone adapts to change in the same way, and not everyone needs or wants the same kind of training. The same goes for organizations, each at a different stage in their AI adoption journey.

At Orium, most of our team are tech folks, so we moved past “AI 101” before we even got to it; our baseline was already further ahead. And since we believe people learn best by doing, we’re moving away from traditional classroom-style training, which is quickly becoming a thing of the past, and instead focusing on enablement that’s closer to the work: micro-learnings, hands-on workshops, coaching, shadowing, and real-time collaboration.

One of our biggest unlocks has simply been having a colleague walk you through how they’ve applied AI in a real project. Peer-to-peer learning has proven far more impactful than any formal training session, with much of our early momentum stemming from early adopters sharing what they’d learned, sparking interest and helping others see what’s possible.

With things moving so fast, our goal is to keep learning relevant, practical, and dynamic, providing support that evolves alongside the tools.

(3) Pay attention to adoption signals → Measure what matters. Rolling out a tool is one thing. Knowing if it’s actually being adopted and making work better is another.

So far, we’ve leaned heavily on qualitative signals like employee feedback, team sentiment, and success stories. That’s been valuable, but as our AI adoption grows, we’re adding more quantitative signals to the mix like usage data, completion rates, and time saved. The goal is to spot patterns early so we can double down on what’s working and quickly fix what’s not.

As we continue to build, we’re also starting to think about managing AI agents in the same way we manage people: by setting expectations, monitoring performance, and making adjustments along the way. If we want employees to collaborate effectively with AI, we have to actively manage those relationships.

As adoption grows, it’s on us to pay attention, course-correct quickly, and ensure AI is meaningfully improving work, not just being used because it can be.

(4) Keep people at the center → Trust, transparency, and human support. Technology doesn’t move a company forward; people do. Adoption sticks when leaders pair leading-edge tech with high-touch, human support at every step.

That starts with strong, two-way communication: understanding how people are feeling, what’s working, what’s not, and where they need more support. And it includes not shying away from the tough questions! The one we hear most often is, “Will AI take my job?” It’s a fair question, and one we’ve tried to address with honesty, sharing what we know when we know it, and being upfront when we don’t have all the answers.

It also means showing up— leaders being present with their teams and demonstrating that they genuinely care. Small actions have an outsized impact: asking how people are doing, helping to alleviate everyday uncertainties, recognizing small wins as they happen, and being open about your own mistakes and learnings. Invite teams into the problem-solving and adoption process so they feel ownership over the change.

We’re fortunate to have a smart, curious, and fast-moving team. They rallied, spoke up, and moved forward together, and because of that, we’re now in an exciting place.

In this AI-powered world, change is constant, and uncertainty feels like the new normal. We have to look ahead and move quickly, so our focus is on making sure our teams are prepared for that reality. We know we won’t get everything right the first time, but we’ll continue to treat AI adoption as a people project first, and a technology project second.

If Orium can remain a place where experimentation is safe, learning is shared, and tools serve the people, not the other way around, we know we can do great things.