Join UsCEO MastermindAI Leadership MastermindExecutive AI ProgramFractional AI Executive1:1 Coaching5-Minute Leader
BooksPower Without PermissionAI Leadership Mastermind
PodcastMeet the Team
ResourcesBlogFree Courses5-Minute Leader AcademyFree AssessmentsResearch and DataTeaching PackagesTestimonialsCEO Coaching ReportJoin UsAdd us on Google

Case Story: Six Production AI Workflows Inside a Multi Location Services Company in Six Months

How a multi location services company went from five disconnected systems to six production AI workflows in six months, with a credential setup that made the rest safe to build on.

By Andreas Pettersson, Founder, Leaders ADAPT

This is one of the two client snapshots on the Fractional AI Executive page, written out in full. Details are generalised to protect the client. The sector is not named.

Quick answer: A multi location services company with a team of about twenty five had sales, service, scheduling and accounting in five disconnected systems. Over a six month fractional engagement we connected Claude to the CRM, email and accounting, built a daily sales briefing and an email drafting workflow for the sales team, turned department interviews into context files, and moved every credential into a managed vault. Six production AI workflows were running by month six.

Company shape

A services company with several locations and a team of about twenty five, two of them senior people working remotely. Sales lived in a CRM, operations in a work management tool, scheduling and accounting in two more, and the five did not talk to each other. Nobody could get a straight answer out of any of them without asking a person to go and look it up. The owner's evenings went to reading updates.

The owner's starting question

How do I get the right information, summarised and accurate, without reading every update myself every night. Behind it sat a second question that most owners of multi location businesses carry: can I step back a day a week without the business slowing down.

What was assessed

Four weeks of discovery. Two hours with the owner to lay out the business and his wishes. Recorded interviews with each leader. Two team members shadowed for a few hours each to see how the work is actually done rather than how it is described. A systems inventory, the repeating patterns across departments, and a scored list of opportunities, presented in a readout.

What was built

  1. Claude connected to the CRM, email and accounting. The first thing the owner wanted was one place to ask a question and get an answer that drew on all three.
  2. A daily sales briefing and an email drafting workflow for the sales team. The briefing runs before the day starts. The drafts go to a person for review before anything leaves the building.
  3. Department by department process interviews turned into context files the AI actually reads. Sales, estimating, operations, HR and service each got a file in the company's own terms, so every answer sounds like the company and not like a generic assistant.
  4. Credentials moved into a managed vault so no password is ever typed into a chat. This came first in practice, because every workflow that followed needed access to a system.

The result after six months

Six production AI workflows running inside the business, on a credential setup that made the rest safe to build on. The leadership team was trained to run the tools before the larger builds, so the workflows have owners inside the company rather than a dependency on us.

What the owner set out to stop doing personally

Reading every project update every night. The target set in discovery was a morning and an evening summary instead. The daily briefing is that summary.

Case story FAQ

What kind of company is this case story about?

A services company with several locations and a team of about twenty five, where sales, service, scheduling and accounting lived in five disconnected systems. The company and its sector are not named, by agreement with the client. The shape of the business is what matters: several locations, several systems, and an owner who could not get a straight answer out of any of them without asking a person to look it up.

What was built in the six months?

Claude connected to the CRM, email and accounting. A daily sales briefing and an email drafting workflow for the sales team. Department by department process interviews turned into context files the AI actually reads. Credentials moved into a managed vault so no password is ever typed into a chat. Six production AI workflows were running inside the business by the end of the six months.

Why did the credential setup come first?

Because every workflow that followed needed access to a system, and the fastest way to lose trust with an owner is to have a password pasted into a chat window. A managed vault meant each workflow could be built and handed over without anyone on the team learning a credential they should not hold. It is unglamorous and it is the reason the rest was safe to build.

Which Leaders ADAPT engagement was this?

The Fractional AI Executive engagement: six months embedded with the leadership team, starting with four weeks of discovery. Details are generalised to protect the client.

Which engagement this was

The Fractional AI Executive engagement. Six months embedded, starting with four weeks of discovery, with the rule that if there is no clear path to return at day 30 we part ways with nothing owed beyond month one. What it costs in the market is on the AI consulting cost page.

If your company runs on five systems that do not talk to each other

The intro call is thirty minutes. Bring the question you cannot get a straight answer to today. We tell you whether a discovery is worth your money, and we say no when it is not.

See the Fractional AI Executive engagement

Details are generalised to protect the client. The six workflows are the same six described on the Fractional AI Executive page.

Read the second case story: a professional services firm.

Leadership insights, straight to your inbox

Practical protocols from a former Canon CEO. No fluff, unsubscribe anytime.

Andreas Pettersson

Andreas Pettersson

Former Canon CEO. Founded and exited Arcules, an AI company backed by Canon and Milestone. Today he coaches CEOs and executives through Leaders ADAPT.

More Posts

Free Leadership Profile & Style Assessments