Small CPA firm practice operations, 21 firms
Summary
Firm-side email, workflow records and channel chat from 21 small US accounting firms: engagement letters and acceptance, busy-season capacity planning, review queues, billing and collections follow-up, software migrations and internal process changes. Return documents, client ledgers, tax figures and any thread about the substance of a client's tax or accounting work are removed before packaging, so the set covers how a small firm runs and holds no client files. Clients appear only as codes.
Collected to order. We collect it from US companies for your order. Scope, volume and timing are confirmed in writing before you sign.
Pricing on request
Request scope and pricingAt a glance
- Planned scope. Volume, sources and timing are confirmed in writing before you sign.
- Records
- 4.5Memails, work items and channel messages
- Size
- 8.3 GB
- Tokens, estimated
- 860M
- Years
- 2018 to 2026
- Firms
- 21
- Owner profile
- 21 US CPA and accounting firms, 5 to 40 staff each, Midwest, Mountain West and Southeast
- Industry
- Accounting and finance
- Data type
- Company operations archive
- Headcount
- 5 to 40 staff per firm
- Region
- United States (Midwest, Mountain West, Southeast)
- Languages
- English
- Formats
- JSONL
- Updated
Sources
| System | Kind | Records |
|---|---|---|
| Microsoft 365 mail | 1.71M messages | |
| Google Workspace mail | 650K messages | |
| Karbon | Projects and wiki | 1.48M work items and comments |
| Microsoft Teams channels | Team chat | 610K messages |
Good for
- training agents to run engagement intake, letters and deadline tracking in a small firm
- modeling busy-season capacity and review bottlenecks from workflow events
- evaluating email drafting for collections and scope-change conversations
- long-horizon reasoning over a firm's process changes across several busy seasons
Fields
| Field | Type | Description |
|---|---|---|
| record_id | string | Stable record id |
| firm_code | string | Code of the contributing firm, for example CF-14 |
| record_type | enum | email, work_item, comment or channel_message |
| created_at | timestamp | Creation time in UTC |
| author_role | enum | partner, manager, senior, staff, admin or client_contact |
| client_ref | string | Client code, present only on engagement and billing records |
| topic | enum | engagement_letter, onboarding, review_queue, billing, collections, firm_admin or technology |
| status_events | array<object> | Workflow status changes with time and role, on work items only |
| text | string | Message, comment or work item text |
Example record
{
"record_id": "CF14-E-0092214",
"firm_code": "CF-14",
"record_type": "email",
"created_at": "2025-02-11T19:06:44Z",
"author_role": "partner",
"client_ref": "[ACCOUNT_41]",
"topic": "engagement_letter",
"status_events": [],
"text": "[PERSON_3], before we start on [ACCOUNT_41] this year the 2024 letter has to change. They added two entities in Q3 and asked for monthly close help that was never in scope. Please split it into two letters, one for the annual compliance work and one for the monthly advisory work, and put both on the new deposit terms. Do not release the year-end workflow until both signed letters are back in Karbon. I will call [PERSON_9] Thursday to walk them through it."
}Licenses available
Exclusive
One buyer. Nobody else can license it during the exclusive term, usually 2 years.
Non-exclusive
Licensed to more than one buyer.
First look
One buyer gets 2 to 4 weeks to make an offer before anyone else sees the dataset.
What ships with it
- Datasheet: what it is, what it is good for, verified size, sources, years, format and limits.
- Provenance file: the owner shown by code, staff notice and consent status, exclusions, mapped to the EU AI Act Article 53 training-data summary and California AB 2013.
- Manifest with a hash for every file.
- QC report and de-identification report.
- Your own copy, with unique canary records.
How it is de-identified
Default exclusions removed, files flattened to text, identifiers found in three layers and replaced with consistent placeholders, and the result measured against a 99% bar for each identifier type in each source.