Problems I have solved
Repetitive operational work that production AI agents now handle for real clients, with the client-confirmed outcome for each and how the system works.
An AI quoting agent for a manufacturer, from days to hours
The quoting inbox took one to three days per quote. An agent now reads the enquiry and the bill of quantities and prices it against the live catalogue, behind a rules engine that keeps every price explainable.
Automating multi-site field reporting, from days to minutes
Daily reporting across many sites took about three days to compile because the data lived in several places. Collector agents now gather it in parallel, a reconciler merges it and a reporter writes the report.
Automating sales order entry from emailed purchase orders
Purchase orders arrive by email as PDFs, spreadsheets and free text, and someone keys every line into the ERP. For one wholesale distributor that became a checked pipeline, then the same approach ran across every country in a group.
Decision support for specialist matching at a global asset manager
Finding the right specialist match took about three days, in a setting with real accountability. An audit-grade agent now returns ranked, cited recommendations, and a person makes the call.
Automating timesheet processing for a weekly payroll run
Hundreds of emailed timesheets a week, a hard payment deadline and no room for a wrong payslip. This is how the reading, checking and reconciliation moved to a pipeline of agents.
Want this in your operation
I build and run production AI agents that take repetitive work off operational teams. Tell me what your team spends too long on.