
Home · Work · Professional services
A professional services firm deployed its first AI workflow within 90 days
A 200-person professional services firm wanted to act on AI but could not set priorities. Armstrat identified and sequenced three high-value automation use cases. The first workflow was fully deployed within 90 days of kickoff.
- Client
- A 200-person professional services firm
- Engagement
- An AI readiness assessment, selection and sequencing of automation use cases, and deployment of the first AI workflow
- Result
- Three high-value automation use cases identified and sequenced, and the first AI workflow fully deployed within 90 days of kickoff
- How we worked
- Assessed readiness inside the firm, narrowed the options to the use cases worth building and stayed with the first one through deployment
Published . Client name withheld.
The situation
A 200-person professional services firm wanted to act on AI. What it lacked was the internal clarity to prioritize. Three questions had no settled answer: which processes to automate, which tools to use and which risks to accept.
With none of those questions settled, every option stayed open, and leadership was paralyzed by the optionality. The firm did not need a longer list of possibilities. It needed a short list it could act on, an order it could defend and a first result it could see.
What made it hard
- There was no shared basis for choosing. Processes, tools and risks were all open at once, and an answer on one shaped the answers on the others.
- Leadership was paralyzed by the optionality. Adding more ideas would have deepened the stall rather than ended it.
- Risk was one of the open questions, so the firm needed to understand what it was taking on before it committed to building anything.
- Progress had to be visible. A plan on its own would not have ended the stall. The firm needed a workflow in use, not just a set of recommendations.

How we worked
Armstrat treated the readiness assessment as the start of delivery, not the end of it. The engagement ran from diagnosis to a working deployment: assess the firm’s readiness, choose the use cases worth building, put them in order, then stay with the first one until it was fully deployed inside the firm.
- 01
Assess readiness before choosing anything
Armstrat started with the firm’s readiness for AI. The assessment worked through the three questions leadership had not been able to settle on its own: which processes were candidates for automation, which tools could serve them and which risks the firm would need to manage.
- 02
Keep only use cases worth building
From the assessment, Armstrat identified the automation use cases worth building and set the rest aside. The test was value to the firm, so the output was a short list of three high-value use cases rather than a long menu of options.
- 03
Put the three in order
Armstrat sequenced the three use cases so leadership faced one decision at a time instead of every decision at once. A clear order replaced the open optionality that had stalled the firm, so everyone knew what came first and what followed.
- 04
Carry the first one into deployment
Armstrat stayed with the work past the plan and into delivery. The first AI workflow was fully deployed within 90 days of kickoff, giving the firm a live result to build on, with the next use cases already in sequence behind it.
The results
Measured against the outcomes agreed at the start.
- The firm’s readiness for AI assessed, with its open questions on processes, tools and risks worked through.
- Three high-value automation use cases identified and sequenced.
- The first AI workflow fully deployed within 90 days of kickoff.
- Leadership moved from paralysis over too many options to a sequenced plan with its first step already live.
What leaders can take from it
Too many options can stall a firm as surely as too few. The fix is rarely more ideas. It is a short list, in a defined order, with clear reasons for what was left off.
A readiness assessment earns its keep by saying no. Naming the use cases worth building also means naming the ones that are not, at least not yet, so leadership attention goes where the value is highest.
Judge AI work by what is live, not by what is planned. Sequencing the three use cases mattered because the first one reached full deployment within 90 days of kickoff.
Questions about this work
How can a professional services firm decide where to start with AI?
Start with a readiness assessment, then choose a small number of use cases and put them in order. In this case a 200-person professional services firm could not decide which processes, tools or risks to prioritize. Armstrat assessed its readiness, identified three high-value automation use cases and sequenced them, and the first AI workflow was fully deployed within 90 days of kickoff.
What should an AI readiness assessment produce?
A short list of use cases worth building, put in order, with the first one carried into deployment. For a 200-person professional services firm, Armstrat’s assessment produced three high-value automation use cases in sequence, and the first AI workflow was fully deployed within 90 days of kickoff. A long, unranked list of ideas would have deepened the paralysis the firm started with.
More work
56.7%Year-on-year revenue growth, from $1.5M to $2.35M56.7% revenue growth for a founder-led agency
$210Krecurring payroll capital released each year, with no service disruptionReleasing $210K a year from an agency cost base
$850Kincremental revenue absorbed in 2023 with only four net hiresAgency absorbs $850K of new revenue with four net hires

Bring us the hard problem.
We agree the outcomes first, then work inside your business until they hold.