NewYour 2027 technology budget is going up. Here's where you still have choices.Read it →

Process Optimization

Working with operations teams to rethink everyday workflows now that AI tools can take on real work. One workflow at a time, with guardrails.

Tell us what's going on

A laptop and an open notebook on a desk
Photo by Bluestonex / Unsplash

Operations teams carry the routine work everything else depends on: finance, HR, grants, contracts, scheduling, reporting, and the hundred small tasks in between. Much of that work is now something AI tools can genuinely help with, from drafting and summarizing to sorting requests and moving information between systems.

But “add AI” isn’t a plan. A tool dropped into a confusing workflow just produces confusion faster. That’s why process work matters again, in a lighter form than it used to take. No 80-page process maps. Instead, find the workflows that eat the most time, look at how they actually run, and redesign them with today’s tools in mind.

Our AI strategy work answers the organization-wide questions: which tools, what policies, how to roll them out. This is the ground-level part, where AI gets put to work in the daily routines of your operations team.

This might be a good fit if:

  • Your operations team spends hours copying information between systems, chasing approvals, or reformatting the same report every month.
  • The same questions land in someone’s inbox every week.
  • Your team has tried AI for a few tasks, but it hasn’t changed how the work actually gets done.
  • You’ve rolled out Copilot, Gemini, Claude, or ChatGPT, and your operations staff still aren’t sure where it fits in their day.
  • A key process only works because one person knows how to do it.
  • You’re growing, and the default answer to more work is another hire.

Tell us what's slowing you down →

How we approach it

Start with the team, not the tool. We sit down with the people who do the work and look at how a process really runs, including the spreadsheets, workarounds, and side conversations. That’s where the time goes, and usually where the opportunity is.

One workflow at a time. We pick the few workflows where a change would make the biggest difference, rather than trying to map everything.

Simplify first, then add tools. Some steps can just go away. Others need clearer ownership. Only then do we look at where AI or automation would genuinely help, and where a person should stay in the loop.

Pilot small and keep what works. Try a change with a few people, see whether it saves time and holds up, then expand. Ideas that don’t work get dropped quickly and cheaply.

Guardrails from the start. Which information can go into which tools, who reviews AI output before it goes anywhere, and how the new way of working gets documented.

Leave your team able to do the next one. The goal is an operations team that can spot the next opportunity and run with it, without calling us.

What it can include

  • Workflow reviews with your operations team: how the work happens now, where it stalls, and what’s worth changing.
  • A short, prioritized list of opportunities for AI and automation, with honest notes on effort, risk, and likely payoff.
  • Pilots of new workflows using the tools you already have, such as Microsoft 365 and Copilot, Google Workspace and Gemini, or Claude or ChatGPT, before buying anything new.
  • Workflow guardrails. For each redesigned workflow: which steps AI handles, who checks its output, and what information it can see.
  • Documentation habits. Templates, owners, and review routines for standard operating procedures (SOPs), so documenting a new workflow becomes part of rolling it out.
  • Technology fit. Whether your systems support the way people work, or force them into workarounds, and what to change.

Why FireOak

We work closely with operations teams in nearly all of our engagements, especially as fractional CIOs. We see where work gets stuck, and we know that operations staff are usually the first to spot what’s broken and the last to get time to fix it.

We also do this in our own operations. We use AI tools every day, and we’re candid about where they help and where they don’t. Here’s how we use them.

And we’re independent. We don’t sell AI tools or automation platforms, so our only interest is what actually works for your team.

Examples from our work

Our own finance reporting. We replaced a series of spreadsheets with financial dashboards we built with Claude. An external financial partner recently told us they were jealous of them.

Building a habit of documentation. Some clients don’t need us to write their procedures. They need their teams to do it, and keep doing it. We’ve worked with organizations to design a model for creating and updating SOPs: templates, ownership, review routines, and the change management that turns documentation into a regular habit. We helped them plan and launch the whole program; their staff wrote the SOPs.

Starting from the goal, not the system. A small organization was working around an old Salesforce setup that nobody knew how to change. Rather than start with “fix Salesforce,” we started with what staff needed to get done. Looking across all of their tools, we found a far simpler way to get there.

Read more client stories →

Common questions

How is this different from your AI strategy work?

AI strategy and implementation is organization-wide: choosing tools, setting policy, rolling them out, and training staff. Process optimization zooms in on specific workflows with the people who run them. AI is often part of the answer, but sometimes the biggest gain comes from removing a step or clarifying who owns something. Many clients do both, in either order.

Our team is skeptical of AI. Is that a problem?

Not at all. Healthy skepticism is useful. We start with small, low-risk pilots, keep people reviewing anything that matters, and let results make the case.

Will you build the automations for us?

Sometimes. Our main role is figuring out what to change and making sure the workflow is ready. Depending on the tools involved, we can help build it, or work alongside your IT provider or the vendor who does.

Do we need to buy new software?

Usually not to start. Most organizations already have more capability than they use, especially in Microsoft 365. If something new would clearly help, we’ll say so.

If the friction is mostly about information that’s hard to find, see our knowledge management work.

Tell us what’s slowing you down

A few sentences is plenty: which work takes too long, what’s frustrating about it, and what you’ve tried. We read every message ourselves and reply within one to two business days.

Prefer email? Write to us at [email protected].

We only use this to reply to you. No mailing lists unless you ask.

Let's keep in touch

Signal, the FireOak newsletter: practical tips and thoughtful takes on technology, knowledge, and AI. No spam. Unsubscribe anytime.