Straight answers, including the ones with a caveat.

A multi-tenant enterprise asset management and maintenance management platform built for oil & gas, for oilfield-services firms, equipment manufacturers and project operators. It covers the equipment master, work orders, inspection and certification, reliability analytics, inventory, projects, workforce and suppliers, with an AI copilot over all of it.

Both, in the sense that matters. It manages the full asset lifecycle the way an EAM does (classification, properties, hierarchy, certification, reliability), and it manages the work performed on those assets the way a CMMS does.

Small and mid-sized businesses in oil & gas that need enterprise-grade asset discipline without an enterprise implementation. The domain model comes from a real oilfield-services maintenance system rather than from a generic asset template.

No. RelyentX is a standalone system of record and runs with no ERP at all. An optional sync with Epicor iScala, SAP or another ERP is part of the committed model and is not yet on the shipped surface.

Rotating equipment, pressure vessels, separators, compressors, generators, flow meters, instruments and lifting gear among others. Classification is a five-level taxonomy and property sets are defined per class, so a class that is not there yet is added rather than forced into a generic shape.

It answers questions from your own data, showing the records it drew on, produces downloadable reports, extracts and draft checklists, and (only when write mode is opted into) proposes allow-listed actions that you approve individually before anything runs.

It retrieves from your own records and is instructed to build answers only from what it finds, to cite those records, and to say so when a fact is missing. It is a language model, so treat what it produces as a draft to check rather than a guarantee, which is why, when an answer draws on retrieved records, those records are shown beside it.

No, and the site will not claim otherwise. RelyentX measures reliability from the events your work produced. There is no forecasting or anomaly-detection model in the product.

Yes. English and Arabic are both first-class, with a full right-to-left layout and an Arabic type face across the product and this site, not a translation layer over a left-to-right interface.

At rest, on EU infrastructure, compute, object storage, the database, the vector and graph stores and outbound email are all pinned to EU regions in the infrastructure code. Two AI services sit outside that pinning and both receive your text: the embedding API that makes records searchable, and the inference API that answers copilot questions. The security page names them.

By the modules you turn on and the people who sign in. A seat covers one named person, and there is no per-site or per-asset charge. We are not publishing the module split or any figures yet; ask and you get both against your own scope.

Roles are drawn from a fixed catalogue and held as an additive set per user, with a view-only floor everyone starts from. Registering and editing equipment, and managing people, suppliers, projects, downloads and acceptance forms each require a named capability at the endpoint; other write routes today require an authenticated member of the tenant. Users are additionally scoped to the tenant’s country, department, project and package layers, and tenant data is read and written through a connection that cannot bypass row-level security. The security page has the detail.

Book a demo. The useful version of that conversation runs against your asset classes and your certificate disciplines, so bring an example of both.