Manufacturing and field service: how a first project here is scoped

What this sector runs on

Manufacturing is the discrete or process production of physical goods. Field service is the maintenance of installed equipment at somebody else's site. They settle in the same ledger: the plant runs on a production plan, a bill of materials and a work order, the service organisation on an install base, a van stock and a dispatch board. Both end at three questions. Was the job done, is it recorded, and can it be invoiced or recovered under warranty? Most software pain here is one of those being answered by a person reading two screens.

Current pressures

What a manufacturing or service director describes as pressure in 2026:

The workflows a project actually touches

A sector does not buy software; a workflow does. Five run underneath almost every project here:

1. Plan to produce. Demand becomes a work order, material is committed, production is confirmed. Scrap, rework and substitution are the traffic nobody designs for.
2. Procure to receive. Purchase order, confirmation, receipt, tolerance matching, invoice. This is the workflow most back-office projects start from, described in ERP and back-office operations.
3. Dispatch to invoice. A service request is triaged, scheduled against a skill and a part, executed on site with evidence captured, signed off, then invoiced.
4. Install base and maintenance history. Serial, configuration, contract cover and past faults; without it, dispatch is guesswork.
5. Warranty and returns. Deciding who pays, and recovering from the supplier before the claim window closes.

When the pressure is inbound volume rather than production, the same triage and routing problem is on customer support automation.

Data and regulatory constraints

• The operations and information technology boundary is real. Integration runs through a historian or broker with one-way flows and an agreed refresh interval. A scope assuming direct access to a controller is mis-scoped.
• Field work is offline first. Technicians work in basements and rural sites, so conflict resolution for a job edited on two devices is a week-one design decision, not a month-five defect.
• Serial and lot genealogy is set by your product, not your software. Machinery, medical equipment and food processing carry different retention regimes, and the strictest one touching your product drives the data model. Netbase JSC holds ISO 27001 certification for information security management and a SOC 2 Type II attestation; neither certifies your equipment or product, and OutsourcingVN names no sector standard for you.
• Technician data is personal data. Location and productivity records about named employees are a legal and works-council question before they are a feature.
• Site evidence becomes contract evidence. Photographs, meter readings and signatures are what a warranty dispute turns on, so retention belongs in the specification.

Published security practice is specific and limited: secure code review and version control, role-based access control, multi-factor authentication on admin dashboards, contributors under non-disclosure agreement, and data-processing agreements on request.

Where AI changes this work, and where it does not

Netbase has delivered anonymised client AI projects including retrieval-based knowledge assistants, document AI and MLOps pipelines. Those map onto this sector: retrieval over service manuals and fault history, document AI over supplier confirmations and delivery notes, and a pipeline that keeps either monitored.

Vision-based quality inspection, predictive maintenance and automated scheduling are scoped separately in discovery, with the data each needs named first. Scheduling is a constraint problem of skills, parts and travel time, not a language problem. The sensor and event pipelines behind any of this are Data and AI Platform Engineering; edge inference for a plant floor or service van is Cloud Platform and DevOps Engineering. Delivery is AI-assisted engineering under human review.

Which route applies to your problem?

Netbase's delivery lifecycle: discovery and strategic alignment; team assembly and architecture planning; agile execution with outcome-based milestones; modular components; training and rollout; ongoing support. On a plant or service network, discovery walks the difficult site, and rollout is booked into a planned shutdown. Three routes follow from the constraint.

• If the constraint is the back-office record — items, suppliers, orders, receipts, movements — start at ERP and back-office operations and scope one workflow, not the system; the ERP procurement and inventory guide frames that buying decision.
• If the constraint is an old system nobody dares change, the commercial route is legacy application modernization: assess first, then move one bounded capability at a time with parity evidence and a tested rollback. The full offer list, and where a quotation starts, is the services page.
• If the constraint is inbound request volume, the queue and routing design is on customer support automation.

Where OutsourcingVN fits, and where it does not

It fits bounded software around the record: the operational data, the job, the evidence, the reporting and the integrations that feed them; modernization of a back-office system that has stopped being upgradeable; and a service application built to your process.

It does not fit control engineering: programmable controllers, supervisory control systems, firmware and safety-rated control stay with your controls vendor. Delivery is remote-first, with Netbase accountable while a team may combine Netbase staff, approved specialists or disclosed partners. No office or capacity outside Hanoi is published. That accountability is set out on the project delivery page.

What related work is published?

Netbase has delivered projects for clients in manufacturing, and, with a local Odoo implementation partner, has consulted on and customized Odoo in Vietnam for manufacturers, an industrial-park purchasing centre and a warranty service network; none of those clients is named. The Odoo implementations record sets out that scope and its limits. Client names, scope and results stay private, so the fit for dispatch, scheduling or on-site inspection is tested in discovery against your own workflow.

A published back-office record shows the transferable mechanics. Since 2020 Netbase has worked as offshore development and managing partner on a multi-tenant cloud ERP SaaS platform for a client that is not named. Its first phase, from 2020 to 2023, covered CRM, real-time messaging, HR, a knowledge base, custom fields and configurable workflows, work and project management, and API integrations, on a React, Laravel and Strapi stack on AWS. A later retail phase covering point of sale, inventory and purchasing was planned work when that profile was written and is not presented as delivered. Other records are on the work page.

A plant manager should discount company size. Netbase's more than 500 clients sit in the United States, Europe and Asia-Pacific, mostly outside Vietnam, as service markets rather than offices, with no project or country counter published; the company began in e-commerce and added consulting and outsourcing in 2016, and its published records are platform work.

So ask for five things instead of a sector logo: a named delivery lead and a named engineer accountable for every change; a discovery output another supplier could use without rework; parity evidence and a rehearsed rollback for anything replacing a running system; a pilot on the difficult site through one full maintenance cycle; and an integration list with a named owner on both sides. How evidence is labelled here is on the methodology page.