Case Studies
What an engagement actually looks like.
Trailhead is new. We don't have signed case studies yet, and we'd rather show you a transparent hypothetical than a vague pitch deck. Below is an illustrative engagement. The firm profile is composite, the workflow is real, and the numbers are pulled from comparable AI deployments we've researched, not from an active client.
We'll replace this page with signed, named case studies as soon as our first engagements wrap and the clients are happy to be named.
Sample · Accounting firm
Cutting month-end close from 4.5 days to 1.8.
The firm
A 12-person mid-market accounting firm in Calgary serving small business clients. Two senior managers, four staff accountants, six support and admin. Annual revenue around $2.5M.
The problem they came with
Month-end close was eating four to five days of senior-manager time every cycle. Most of it was repetitive: pulling data from QuickBooks Online, formatting into client-facing PDF packages, drafting commentary on variances, and chasing missing receipts. The partners had each lost a full week per month to work that was technically delegatable but practically couldn't be. No one else in the firm could draft commentary that read like a partner.
The Readiness Audit (1 week)
- Half-day workshop with the two managing partners and the senior accountant who runs close logistics.
- Process-mapped the 12-step close workflow. Six steps were judged AI-eligible.
- Ranked use-case shortlist: month-end close commentary (highest ROI), receipt OCR and categorisation, recurring client communications.
- Recommended tooling: Claude for Teams plus a custom integration into QBO. Estimated 32–40 partner hours saved per month if Sprint #1 hit its KPI.
- Deliverable: 18-page PDF report + 60-minute readout to partners.
Audit fee: $4,500 (fixed). Identified annual savings: ~$57,000 in partner time at standard charge-out rates. Comfortably past the 5× guarantee.
Implementation Sprint #1 (4 weeks)
- Built a custom integration that pulls trial balance, P&L, and balance sheet from QBO at month-end on a schedule.
- Wrote a Claude Skill that runs variance analysis prompts against the prior 12 months of the same client's data and drafts management commentary in the firm's house voice (trained on 30 prior partner-written packages).
- Deployed inside the firm's existing Microsoft 365 workspace, so partners review and edit in Word as before, but the draft starts at 80% rather than 0%.
- Captured a baseline KPI before launch (average close time across last six months) and a 30-day post-launch KPI.
- Delivered: working integration, a runbook, training session, and a fallback procedure for when QBO data is incomplete.
Sprint fee: $14,000 (fixed).
The result
- Baseline: 4.5 days average close time across 12 monthly clients.
- After 90 days: 1.8 days average close time across the same client base.
- Senior hours recovered: ~32 hours per month, valued at standard charge-out of $150/hr = ~$4,800/month of recoverable capacity.
- Payback period for the audit + Sprint #1: just under 4 months.
What came next
The firm signed a 6-month Fractional AI Architect retainer to tackle Sprint #2 (receipt OCR + categorisation), keep the close workflow tuned, and roll out a similar pattern to two other recurring deliverables. This is the most common shape an engagement takes: Audit → Sprint #1 → retainer if it worked.
This is what a typical entry looks like: one audit, one sprint, then decide. Most firms either retain us for ongoing work or take the runbook and run it themselves.
Same pattern, your firm
We'll model the math for your firm in 30 minutes.
Book a free AI Readiness Conversation. We come prepared with three workflows specific to your firm and a rough projection of what AI would be worth in recovered time.