AI for CPA and accounting firms. More capacity before busy season.
AI can chase missing client documents, sort and extract what arrives, prepare workpapers and reconciliations for review, draft engagement letters and client emails and search the firm's past work. Modern Thesis helps managing partners of CPA and accounting firms put that to work, so the firm takes on more clients without hiring staff it can't find.
At a glance
- Who it's for: managing partners, partners and COOs of CPA and accounting firms with $10M+ in annual revenue, including tax, audit, bookkeeping and advisory practices.
- What Modern Thesis does: AI strategy, change management, implementation and ongoing stewardship, from one team.
- How it starts: a strategic assessment that measures how the work moves today and ranks where AI pays first. See the assessment.
- Where: based in the Denver area, with an office at 11001 W 120th Ave, Ste 400, Broomfield, CO 80021. On site across the Denver metro and Colorado, and remote anywhere in the US.
- What it doesn't do: give tax or legal advice, sign or review returns or audits for you, or publish a rate card. Each engagement is priced to its scope after a first conversation.
Who is this page for?
This page is for the people who run a CPA or accounting firm. The pattern is familiar: busy season compresses the year, client documents arrive late and incomplete, staff spend hours chasing and sorting before any real work starts, and hiring experienced accountants is hard. Many firms are also trying to grow advisory work, which needs partner time that tax season consumes. That's a capacity problem, and AI addresses capacity.
Where does AI help an accounting firm?
These are examples of what can be built, not past results:
- Send each client a personalized list of what's still missing and follow up until it arrives.
- Sort uploaded documents by type and year and extract the figures into the firm's workpapers for preparer review.
- Prepare bank and account reconciliations for bookkeeping clients and flag transactions that need a person.
- Draft engagement letters, client emails and notices responses from the firm's templates.
- Search the firm's past memos, workpapers and research in plain language, with the source of every answer.
How should a firm handle client data and AI?
Carefully, and with the rules of the profession in mind. Section 7216 of the Internal Revenue Code limits how tax return preparers use and disclose taxpayer information, and the IRS publishes guidance on it.1 A firm system should keep client data private, agreed in writing, never used to train AI models and limited to the work. Modern Thesis builds to those rules. We don't give tax or legal advice; your firm decides what its obligations require, including any client consents.
What does Modern Thesis do for a firm like this?
Modern Thesis covers four disciplines with one team. In strategy, Jim Coleman and the team map how the work moves from first contact to final invoice. In people, we work with the professionals and staff on incentives and training, so they trust and use what gets built. In implementation, our AI specialists and Forward Deployed Engineers build private systems on the firm's own documents and connect them to the software you run. In stewardship, we stay to check accuracy, keep pace with new AI models and report the return against the baseline.
How does an engagement start?
Every engagement starts with a strategic assessment. We sit with the people who do the work, map how it moves through the firm today, and measure a baseline on the workflows that matter most: hours per return or engagement, days waiting on client documents, review time, write-downs and partner hours available for advisory. Then we rank the opportunities by return and define the first build. If AI won't pay for itself somewhere in your business, we'll tell you.
When is something else the better choice?
If your practice management or tax software's AI features already handle document collection well, use them. If your firm is small and mostly individual returns, a client portal with reminders may be enough. If the main need is better writing, the ChatGPT or Copilot licenses you already pay for may be enough; our comparison of Copilot or ChatGPT and an AI partner walks through the difference.
How Modern Thesis handles this
We put AI to work inside xFusion, the 70-person company Jim co-founded, and then for several of its clients. For an accounting firm, we usually start with document collection and preparation before busy season, because the payoff arrives in the same year. Ongoing stewardship has no set end date, and you own your data, accounts, code and documentation throughout.
Questions
How are CPA firms using AI?
Mostly on the work before and around the professional judgment: collecting and sorting documents, extracting figures, preparing reconciliations, drafting letters and searching past work. The preparer and reviewer stay responsible. The return shows up as more engagements per person and a less painful busy season.
Does AI replace our tax or audit software?
No. Keep it. Your tax, audit and practice management software stays at the center. AI works around it: getting cleaner inputs in, drafting the communication and finding answers in your past work. Where the vendor allows, it can read from and write to those systems.
Can AI help an accounting firm grow advisory services?
Indirectly, by freeing partner and manager time. Advisory work needs senior attention, and compliance work tends to absorb it. When document chasing, preparation and review prep take less time, that time can go to clients who want more than a return. AI can also draft first versions of advisory reports from the client's numbers for review.