A lot of businesses are leaning on their IT teams to lead AI adoption across the organisation.
It's an understandable place to start. There are platforms to assess, systems to connect and decisions to make about security and business information. IT has a big part to play in getting that right.
I'd start with what the business is trying to achieve. A team might want to support more customers, improve a service or find room to grow without adding pressure to people who are already stretched.
For me, the useful conversation is how IT and operational leaders work towards those outcomes together. The people accountable for the work should help decide how it changes. IT can help shape those decisions from the beginning, including what AI might make possible.
That means looking at your people and processes, how work actually gets done, alongside the tools and systems they use. Then there's the company brain: The business knowledge and ways of working, organised so your people and AI can draw on them.
Making a tool available gives people somewhere to start. From there, who helps the team explore how to use it? Who can change the process and give people time to learn? How will they judge whether the change is useful?
The answers will vary between businesses. An IT leader may already be closely involved in how the business runs. Elsewhere, those responsibilities sit across several departments. A shared plan helps people understand what they're contributing and who will make the decisions as the work develops.
Combining Technical and Operational Expertise
Your tools and systems are part of that picture. Information needs to move between applications, access needs to be controlled and someone needs to keep the connections working.
If AI can read information from your CRM or finance platform, the business needs to agree what it should do with it. If it can update a record, who checks the change? What happens when information is missing or the answer is wrong?
IT brings expertise to those decisions. The people responsible for the work understand what should happen and where a mistake could cause problems. Working through an actual task together helps both groups see the possibilities and the checks it needs.
An email draft might be a useful first application. Looking at the process around it could reveal another opportunity: Perhaps someone spends much longer gathering information or waiting for a handover than they do writing the email.
That gives the team a more useful conversation about where AI could help. The department can explain how the work happens today, while IT can suggest ways to connect information or make a different approach practical.
Those conversations can change the original brief. I'd make room for them before committing to a particular tool or solution.
Starting With the Work and the People Doing It
I'd start with a part of the business where the team sees room to improve its service or capacity.
If customer onboarding ties up your most experienced people, what takes their time? Are they finding information, correcting incomplete handovers or answering questions that come up with every new customer?
That gives the team something specific to explore with IT. It also provides a way to compare the current process with a different approach.
Take a handover from sales into delivery. Connecting the CRM to the delivery tools helps move information. The teams can then agree what a complete handover contains, which commitments need checking and who sorts out anything missing.
This is where the company brain becomes useful. The agreed process and examples of a good handover give AI the context to work from, alongside the current customer information in your systems. A reusable agent skill can turn that knowledge into steps and checks for preparing the handover.
The people doing the work can help build and test it, with IT supporting access and connections. Together, they can agree who reviews the output and who has responsibility for improving the process as they learn.
Keeping that knowledge current is part of the work too. The team can agree who updates it as the process changes.
That's where the growth opportunity sits. If experienced people can support more customers without a growing backlog, the business has more capacity. It helps to agree how the team will use it, whether that's improving service or taking on work it previously couldn't.
Time and Support for Department Heads
Department heads know their people and where the work gets stuck. That's a useful place to start.
What we're seeing in some organisations is those leaders being given an AI subscription and asked to explore it alongside their existing responsibilities. There's something to learn from that, but they've also got a department to run. Giving them some room and support makes it a much fairer test.
They need access to the tools and room to learn. Technical support helps them put ideas into practice, while guidance on what's possible can help them choose where to start.
That guidance could come from a colleague who has already applied AI to similar work, or from someone outside the business. The department head brings the context; the person supporting them can help connect that context to a practical use of AI.
I'd pick a recurring task and establish how it works today. Give the team time to test a different approach, then compare the effort involved and the quality of the result. Pay attention to where a person's judgement helps, and use what the team learns to shape the next step.
I can understand why a business initially sees AI as another software subscription. Looking at the work gives it a way to assess the wider investment, including preparation and support. Can the team serve more customers or take on work it couldn't previously manage? Those are useful questions to keep coming back to.
AI Leadership That Fits the Business
If you're looking for one person who understands AI, knows how a business runs and can bring people along with them, you're almost looking for a unicorn. Those people exist, and for some businesses a dedicated hire will make sense.
You can also do both. Have someone inside the business own it, whether they're already in the team or a new hire, and give them technical and advisory support from people who've done this work before.
With that support, they can explore what's possible and get the implementation moving. The aim is to reach useful results sooner while building capability in their own team.
At ELab, that's the role we can play: Helping your team connect the dots. We start with your people and processes, how work actually gets done, then help build the company brain around that knowledge.
We connect it with the tools and systems your team uses, and build and test the workflows and agent skills together. Then we support people as they put them into practice.
The budget can go towards a hire, outside support or a mix of both. I'd look at what each arrangement could get done for the business, including implementation and the support your team will need over time.
Whichever arrangement you choose, I'd agree who owns the business outcome and which decisions they can make. Give that person the time and support to work with IT and the department heads.
Then look at what's changed in the work. Can the team take on more customers? Are handovers more complete? Has rework reduced?
That's what I'd keep coming back to: What your people can now do, and the capacity that creates for the business to grow.