Business practice
Advice that ships.
A consulting practice for accounting firms, law practices, nonprofits, and other service organizations that have been handed enough slide decks. Every engagement ends with something running, measured, and owned by your team.
The situation
None of this is hypothetical.
The work is invisible.
Nobody in your firm can say, precisely, how the highest-cost workflow actually runs today. It lives in habits, inboxes, and two people who must never take vacation at the same time. What is not written down cannot be improved, priced, or handed off.
AI arrived without permission.
Your team is already using AI tools. Quietly, individually, and on client data. Nearly half of professional services firms have had confidential information pasted into public tools. The exposure exists whether or not anyone has asked about it, and your insurer has started asking.
The last initiative did not stick.
You have bought software, hired consultants, or run a pilot before. It produced a document and then nothing changed. This is the industry norm: the overwhelming majority of AI pilots produce no measurable financial impact. The problem was never the tool. It was that nothing shipped.
What we do
Our services.
For organizations trying to operate in a landscape that keeps moving, you need guidance that understands where things stand today, reads where they are heading, and knows how to manage that change without adding risk. You need someone who has run the work rather than only studied it, and who can bring real solutions to the problems sitting in front of you now.
We work across service firms and nonprofits, which shows us where these problems repeat and where they genuinely differ, balanced against what it actually takes to run your organization on an ordinary Tuesday. An engaged team is what makes any of this hold, so people are part of the plan from the first week rather than a training step at the end. And as experienced staff retire and institutional knowledge walks out with them, getting that knowledge written down and owned stops being a nice-to-have.
AI governance and adoption
Shadow AI Exposure
You have not asked your team what they are using, because you are not sure you want the answer. Two weeks from now you know, you have a short list you can act on, and the uncertainty is gone.
- Anonymous team survey that gets you the real picture
- Which client data has moved into which tool
- Three things to do this week at no cost
AI Use Policy
The renewal form asks whether you have a written AI policy and there is no box for "we are working on it." Within weeks the answer is yes, and it holds up when someone reads past the first page.
- Tools tiered, data classes defined in plain language
- Signed by every member of staff, and filed
- A named owner and a review date, so it stays true
AI Adoption
Buying the tool was the easy part. Getting people to use it on real work is where most rollouts stall. You end up with a defined set of use cases, the people who own them, and adoption you can see in the work rather than in a licence count.
- The work AI should touch, and the work it should not
- Named use cases, each with an owner and a measure
- Adoption tracked where the work happens
AI Change Management
A rollout that arrives as an announcement gets treated as one. You get a change plan built for the people doing the work, so the new way survives the first busy week instead of quietly reverting to the old one.
- Where the resistance actually sits, asked rather than assumed
- Training built around the real task, not the feature list
- A checkpoint after go-live, when the drift usually starts
Retention and process
Client Retention
Clients and members do not announce that they are leaving. They go quiet. You reach a position where the signal arrives early enough to act on, and the handoffs that used to cost you relationships are owned.
- The exit points found from the accounts you already lost
- First ninety days rebuilt around visible value
- A review cadence that surfaces a cooling account
Process and Workflow
The work gets done because specific people remember how. You end up with the workflows that actually cost you money written, owned, and measured, so the organization stops being one resignation away from a problem.
- Your highest-cost workflow mapped as it truly runs
- A cost baseline every later claim is measured against
- Procedure written in your team's own words
Customer Journey Mapping
Nobody has drawn the path a client or member actually takes through your organization, so every team optimizes its own piece. You get the whole path mapped end to end, with the handoffs that cost you named and owned.
- Every touchpoint from first contact through renewal
- The handoffs where people go quiet, marked and owned
- One view the whole team works from
Onboarding and Offboarding
The first ninety days set the whole relationship, and the exit tells you what to fix. You get both designed on purpose, instead of improvised by whoever happened to pick up the account.
- A first ninety days built around visible value
- A structured exit that returns a reason, not a guess
- The same standard whoever happens to run it
Operations
SOP and Documentation
The procedures that matter live in habit and in two people's heads. You end up with the ones worth writing captured in your team's own words, owned by a name, and reviewed on a date so they stay true.
- The workflows worth writing, chosen by what they cost
- Procedure your team follows because your team wrote it
- An owner and a review date on every document
Department Structure
Roles grew around the people who happened to be there. You reach a structure where responsibility is clear, decisions have an owner, and nothing important sits in the gap between two job titles.
- Who owns what, written down and agreed out loud
- The decisions that queue, and where they should sit
- Gaps and overlaps surfaced before they cost you
Headcount Planning
The instinct when work piles up is to hire, and sometimes that is right. You get an honest read on which pressure is a staffing problem and which is a process problem, before the offer goes out.
- Current capacity measured rather than estimated
- Which load a fix absorbs, and which needs a person
- A plan you can defend to a board or a partner group
Training and Enablement
New people learn by shadowing whoever is free, so the standard drifts with every hire. You end up with a path that gets someone productive on a schedule you can actually predict.
- A ramp with milestones instead of a shadowing rota
- Built on the procedures your team already follows
- Refreshers aimed at where the work actually slips
How we engage
The 30-Day Proof
You have paid for advice before and received a document. Thirty days from now one workflow runs differently, a named person owns it, and you have a before and after number instead of an impression.
- One fix built and live before the engagement ends
- Every finding scored on impact and complexity
- Nothing running on day thirty means no invoice
Fractional Operations
The operation is running on you, and everyone knows it. You reach a point where it has an owner who is not you, decisions stop queueing behind your calendar, and you still make the calls that matter.
- A weekly operating cadence that is run, not attended
- Every open initiative governed, sequenced and owned
- One client at a time, with a successor built in
Results
8
Industries operated in, not just advised
20%
Lower support cost, one year
50%
Faster client onboarding
Results we have been able to achieve in prior operating roles. Every firm and every industry is different, so treat these as evidence of the approach, not a forecast of your numbers.
Fit
This works when it fits.
Works well when
- You are a professional services firm of roughly fifteen to one hundred fifty people.
- The owner or managing partner is in the room and can make a decision.
- You bill for time, so hours recovered convert directly into dollars.
- You suspect your team is already using AI and would rather know than wonder.
- You want one thing measurably improved more than you want a strategy document.
Works badly when
- You have already decided the answer is building a custom AI agent.
- Your core problem is pipeline or marketing, because that is not this lane.
- You are under ten people, where the pain is real but the budget is not.
- You are an enterprise with a procurement cycle longer than the engagement.
- Nobody with authority can give the work four scheduled hours in thirty days.
Questions
Asked before, answered plainly.
Exactly what it says. If day thirty arrives and no fix is running and measured in your business, the invoice is zero. Not discounted. Zero. The guarantee is safe for us to make because it attaches to shipping, which we control, not to a promised percentage, which nobody honestly controls. If the fix breaks within sixty days for a reason inside scope, we repair it at no charge.
No. It is an operations practice with an AI governance wedge. Most findings in a typical diagnostic are not AI at all. They are unwritten processes, duplicate software, and broken handoffs. AI is one disposition on the scoring matrix, and it has to earn its place there like everything else.
The engagement is structured so that a document cannot be the final deliverable. The readout includes a fix that is already live, its written procedure, and its measured before-and-after. Findings are included, but findings are the beginning of the engagement, not the end of it.
No, and we say so on the first call and in the document itself. It is an operational policy: which tools are sanctioned, what data may enter them, and how staff acknowledge it. It is the document your insurer and auditor ask about. Your attorney should review it, and it is written so that review is fast.
Four scheduled hours of access across the thirty days, timely decisions on the fix we scope together, and honesty about how the work really happens today. The guarantee is a two-way agreement. We commit to shipping. You commit to being reachable enough that shipping is possible.