- Complexity: Standard
- Deployment: Platform-native
The problem
Field teams on construction sites record concrete deliveries, slump tests and cube test results by hand, then re-key that information into spreadsheets or registers for QA, sustainability and finance to use later. The lag between a delivery and its recorded result creates rework and slows reconciliation with supplier invoices.
- Site staff manually transcribe docket numbers, slump results and pour locations
- Cube test results due at seven and twenty eight days are tracked by memory or spreadsheet reminders
- Delivery invoices are reconciled separately from the field record, creating mismatches
- QA, sustainability and finance each need the same delivery data re-entered for their own systems
The solution
A delivery tracking workflow turns two photos sent over a messaging app into a complete structured record that feeds QA, sustainability and finance, while a site engineer confirms the results that matter.
- Receives docket and slump test photos submitted by field staff over WhatsApp
- Extracts docket number, slump result, pour location and mix ID from the images
- Writes the structured record to the SharePoint delivery register
- Generates the lab request form for the cube test samples
- Tracks pending seven day and twenty eight day cube results and sends reminders until they are logged
- Reconciles the completed record against the delivery invoice
Where people stay in control
- A site engineer confirms lab results before the record is treated as final
- QA, sustainability and finance draw from the same confirmed record rather than separate manual entries
- Reminders for overdue cube results go to a named person, not an automatic escalation
Systems and data
The workflow reads photos submitted through WhatsApp Business API, writes structured delivery, slump and cube test records to a SharePoint register, and reconciles those records against delivery invoices.
- WhatsApp Business API
- SharePoint
Results
- Inspection report turnaround was 30 to 50 percent faster in the documented deployment
- Teams saved 8 to 14 hours of manual data entry per week per team in the documented deployment
Why platform-native
The workflow reads a single messaging channel and writes to one shared register with limited data sensitivity, so it runs platform-native, directly on the model platform with Altivus providing monitoring and governance.
Frequently asked questions
Does this replace the site team's data entry entirely?
It removes manual re-entry of docket and test data, since the agent builds the structured record directly from photos, but a site engineer still confirms lab results.
What messaging channel does the workflow use?
Field staff send photos through WhatsApp, and the agent extracts the structured fields from the images.
Does it track lab results automatically?
Yes. It tracks pending seven day and twenty eight day cube test results and sends reminders until they are logged.