Concrete and sitework foreman daily reports in English and Spanish, 10 firms
Summary
Foreman daily reports, pour and equipment logs, crew channel messages and project mail from 10 US concrete and civil site work subcontractors. About 40% of the daily reports are written in Spanish or switch languages mid-sentence, with real trade vocabulary for formwork, rebar, compaction and underground utilities. Crew members, inspectors, sites and general contractors are replaced with consistent placeholders, and crew sizes are kept as counts only.
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
- 3.6Memails, chat messages, daily reports and log entries
- Size
- 6.8 GB
- Tokens, estimated
- 710M
- Years
- 2016 to 2026
- Firms
- 10
- Owner profile
- 10 US concrete and civil site work subcontractors, 25 to 160 staff each, Texas, Southwest and Southeast
- Industry
- Construction
- Data type
- Field operations
- Headcount
- 25 to 160 staff per firm
- Region
- United States (Texas, Southwest, Southeast)
- Languages
- English, Spanish
- Formats
- JSONL, Parquet
- Updated
Sources
| System | Kind | Records |
|---|---|---|
| Procore daily logs | Field operations | 212K daily reports |
| Shared drives (pour logs, equipment logs) | Shared files | 204K log entries |
| Slack public channels | Team chat | 1.24M messages |
| Google Workspace mail | 1.9M messages |
Good for
- evaluating Spanish and code-switched text understanding in a technical trade
- extracting crew size, quantities, delays and inspections from daily reports
- training agents that build delay and weather narratives for schedule updates
- translating field reports into English summaries for project managers
Fields
| Field | Type | Description |
|---|---|---|
| report_id | string | Unique daily report id |
| firm_code | string | Placeholder for the subcontractor |
| project_id | string | Placeholder for the job |
| report_date | timestamp | When the foreman submitted the report |
| foreman_id | string | Placeholder for the foreman, stable across reports |
| language | enum | en, es or mixed |
| crew_count | int | Workers on site that day |
| work_completed | string | Work narrative as written |
| quantities | array<string> | Installed quantities with units |
| delays | string | Delays and their causes |
| weather | string | Weather as recorded by the foreman |
| inspections | array<string> | Inspections and tests on site that day |
Example record
{
"report_id": "CS10-DR-0117462",
"firm_code": "[COMPANY_6]",
"project_id": "[PROJECT_903]",
"report_date": "2024-09-12T17:20:00-05:00",
"foreman_id": "[PERSON_215]",
"language": "mixed",
"crew_count": 14,
"work_completed": "Colamos losa nivel 2 área B, 186 CY de 5000 psi con bomba de 47 m. Acabado con llana mecánica, curing compound a las 3:40. Empezamos cimbra de columnas línea 6 a 8.",
"quantities": [
"186 CY placed",
"9,400 SF slab finished",
"12 columns formed"
],
"delays": "Bomba llegó 50 min tarde. 2 trucks rejected, slump 6.5 in, spec max 5 in.",
"weather": "92F, humedad alta, sin lluvia",
"inspections": [
"Special inspector [PERSON_17] cast 4 sets of cylinders",
"City inspector passed rebar on level 3 deck"
]
}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.