An owner-led oilfield services contractor, part of a family-owned group, had years of field, safety, project, and financial data spread across separate systems. nBrain connected them into one private Brain, then built the tools its estimators and billing team asked for. This story is still being written, and we say so.
Client names available upon request.
Each division ran its own fleet, telematics, safety, project, spend, accounting, and payroll systems. A leadership question that crossed two of them meant manual exports and a spreadsheet. Years of meetings and email were effectively unsearchable.
Two jobs took the most hands. Unit-price bids for major operators were built by hand from crew-day estimates. And every field ticket had to be checked against GPS, safety forms, and timecards before it could be invoiced through each customer's portal.
Fleet, GPS, safety, projects, spend, accounting, payroll, the ERP, meetings, email, and documents, with access by company and role and a source on every answer.
Talked with staff by voice or text about where their time goes, and turned it into a ranked list of use cases.
Drafts every line of an operator's price book with a confidence score, and sends uncertain lines to an estimator review queue.
Matches each ticket against GPS, safety forms, and payroll, routes exceptions to a queue, and bundles clean tickets for invoicing.
Agent memory, personalization, and actions shipped switched off by default, each waiting for a person to approve it.
The engagement started with one agreement covering every company in the group. The data work went fast. Turning it into daily use went slower, because each company had its own systems, its own priorities, and in some cases a core system being replaced mid-build.
The fix was focus: one business, one named owner, one or two workflows with a written definition of done, tested with the people who do the work. That is how the bid calculator reached live use, and it is how billing is moving to production.
An expert estimator reviewed the bid calculator, accepted it, and used it on live work. Field-ticket billing checks are now in testing with staff on live ERP data.
Lines the model is confident about go to bulk approval; the rest wait for an estimator. Every edit is shared across the team.
Exceptions to a queue, clean tickets bundled for invoicing, and hours returned to the billing team measured against a baseline.
Customer, staff, product, and program details have been removed to protect the client's confidentiality. Client names available upon request.
In one strategy session we map where a Brain would return the most time, what we'd build first, and how we'd measure it from month one.
Prefer email? hello@nbrain.ai