Custom software agencies: client email, chat and delivery records, 14 firms
Summary
The working records of 14 small US software agencies that build web and mobile apps for clients: client and internal email, public Slack channels, Jira boards and shared project documents. It covers the whole delivery cycle, from discovery and estimates to change requests, QA rounds, launch and support retainers. Client names, contact details, credentials and staff identities are replaced with consistent placeholders. Client source code is not included because the clients own it.
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
- 25.7Memails and chat messages
- Size
- 64 GB
- Tokens, estimated
- 6.4B
- Years
- 2015 to 2026
- Firms
- 14
- Owner profile
- 14 US custom software and app development agencies, 6 to 45 staff each, several with nearshore developers in Latin America
- Industry
- Software
- Data type
- Company operations archive
- Headcount
- 6 to 45 staff per firm
- Region
- United States
- Languages
- English, Spanish
- Formats
- JSONL
- Updated
Sources
| System | Kind | Records |
|---|---|---|
| Google Workspace mail | 12.9M emails | |
| Microsoft 365 mail | 3.9M emails | |
| Slack public channels | Team chat | 8.9M messages |
| Jira | Projects and wiki | 1.2M issues |
| Google Drive | Shared files | 184K proposals, specs and QA documents |
Good for
- training project agents to turn client email into scoped tickets and change requests
- evaluating estimates and timelines against what each project actually took
- learning client communication across discovery, delivery, QA and launch
- testing bilingual coordination between US leads and nearshore developers
Fields
| Field | Type | Description |
|---|---|---|
| record_id | string | Stable de-identified record id |
| firm_code | string | Pool member code |
| source | enum | email, chat, issue or document |
| thread_id | string | Email thread or chat thread id |
| sent_at | timestamp | Send or post time |
| sender_role | enum | client, project manager, developer, designer, QA or executive |
| recipients | array<string> | De-identified recipients |
| subject | string | Email subject or channel name |
| body | string | Text with quoted replies and signatures stripped |
| folder_path | string | Original mailbox folder or channel |
| project_code | string | De-identified client project |
Example record
{
"record_id": "E-F07-0093318",
"firm_code": "F07",
"source": "email",
"thread_id": "TH-F07-221904",
"sent_at": "2022-06-14T15:22:10Z",
"sender_role": "client",
"recipients": [
"[PERSON_3]",
"[PERSON_18]"
],
"subject": "RE: Sprint 9 demo follow-ups",
"body": "Thanks for the demo. Two things before we sign off. The driver app still logs people out when they switch to the camera, and our ops team needs the CSV export to include the stop notes column. Is the export a change request or part of the original scope? We need it before the [CITY_2] pilot on the 28th.",
"folder_path": "Inbox/Clients/[COMPANY_11]",
"project_code": "P-F07-031"
}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.