Data platform readiness: check ownership and quality before you build

OutsourcingVN is operated by Netbase JSC, which scopes and builds data and workflow systems, so we have an interest in your decision. The checks apply whoever builds the platform. This guide is for CTOs, data and operations owners, and finance leads asked to fund a warehouse, reporting layer or AI initiative.

What this guide covers

• Which decision is the platform for?
• Who owns each dataset?
• How good is the data, measured how?
• Can the platform get the data out, safely?
• How do you run the assessment?
• A worked scenario: a distributor that wants margin reporting
• What has Netbase delivered around business data?
• Which questions should you ask a supplier?
• What failure modes should you watch for?
• Common questions
• How this guide is sourced
• Bring your decisions and a data sample

Which decision is the platform for?

Start with three to five decisions the business wants to make faster or with more confidence: which products to reorder, which customers are at risk, where margin leaks between order and invoice. Each decision names the data it needs, how fresh that data must be, and who acts on the answer.

Without that list, a platform project collects everything it can reach and then struggles to show value. With it, the assessment has a boundary. A dataset that no listed decision uses can wait for a later phase, however interesting it looks.

Who owns each dataset?

Ownership is the first gap most assessments find. For every dataset in scope, record the system of record, the business owner who decides what a field means, the technical owner who keeps the feed running, and what happens when two systems disagree.

Customer records are the usual test case. Sales, support, billing and the web store each hold a version. If nobody has decided which one wins for address, status or credit terms, a platform will faithfully reproduce the conflict in every report. Deciding that rule is business work, not engineering work, and no tool makes the decision for you.

How good is the data, measured how?

The UK Government Data Quality Framework, published in December 2020, uses six dimensions adopted from DAMA UK: completeness, uniqueness, consistency, timeliness, validity and accuracy. They make a practical checklist because each can be profiled on a sample before any build.

Profile first and set thresholds per decision. A demand forecast can tolerate some missing categories; an invoice reconciliation cannot tolerate duplicate customers.

Can the platform get the data out, safely?

For AI use cases, the NIST AI Risk Management Framework (released January 2023) groups its practices into four functions: govern, map, measure and manage. Readiness work sits mostly in map and govern: knowing which data feeds a model or retrieval index, who approved it, and how it will be checked. The RAG context engineering guide covers source registers for retrieval in detail.

A worked scenario: a distributor that wants margin reporting

A distributor wants weekly margin by customer and product line. Orders sit in an ERP, customer discount agreements in spreadsheets, freight costs in a carrier portal, and returns in a support desk. The scenario is illustrative and describes no client.

Profiling shows many older order lines with no product category, customers duplicated between the ERP and the support desk, and freight costs available only as monthly PDF invoices. The assessment ranks the gaps: agree the customer master in the ERP, backfill categories for the top-selling lines, and request a freight export from the carrier. The spreadsheet discount agreements stay out of phase one because no owner will commit to maintaining it.

Phase one delivers margin before freight for the top product lines, reconciled to finance totals each week. Freight joins in phase two once the export exists. The distributor gets a trustworthy partial answer in weeks instead of an unreliable complete one in months. Where the gaps sit in order capture or back-office processes, the ERP and back-office operations and order and payment operations workflows show where a system fix belongs, and document and approval workflows covers data that arrives on paper or PDF.

What has Netbase delivered around business data?

• ERP work, delivered with a local partner. Working with a local Odoo implementation partner, Netbase has consulted on and customized Odoo in Vietnam for retail store chains, a food and beverage franchise chain, manufacturers, trading and distribution companies and other organisations that are not named.
• ERP integration and upgrade experience. Netbase's ERP experience covers e-commerce integration with ERP by API synchronisation of product catalogues, inventory, orders and customer data, and a Vtiger CRM to ERP upgrade, with modules including finance and accounting, inventory, sales orders, CRM and reporting.
• A long-running SaaS platform. Since 2020 Netbase has worked as offshore development and managing partner on a multi-tenant cloud ERP SaaS for a US client that is not named. Its retail phase 2 was planned work when the profile was written and is not presented as delivered.
• Anonymised AI delivery. Netbase has delivered anonymised client AI projects including retrieval-based knowledge assistants, document AI and MLOps pipelines.

Netbase's compliance practices are GDPR alignment for data privacy in Europe, HIPAA-aligned methodologies for healthcare data handling, and CCPA compliance for clients with U.S. customer bases; these are practices, and a client's own legal position stays with its advisers. The data security guide covers contract and access controls. Security practices include secure code review and version control, role-based access control, MFA for admin dashboards, contributors under NDA, and NDAs and DPAs on request.

Which questions should you ask a supplier?

What failure modes should you watch for?

How this guide is sourced

Statements about Netbase map to approved claims backed by attested company facts listed under Sources, in their registered wording. The quality dimensions come from the UK Government Data Quality Framework and the four functions from the NIST AI Risk Management Framework, both dated below. The distributor scenario and its figures are illustrative. Netbase has no published data-warehouse or analytics-platform delivery record: the evidence here is ERP, integration, SaaS and anonymised AI work, and no measured client outcome is published. See the methodology for how claims are reviewed.

Bring your decisions and a data sample

A first conversation goes furthest with the decisions you want to support, a list of source systems and owners, and a sample extract. Submit a project with them. When the workflow itself is still unclear, an AI Workflow Blueprint is the assessment route, and the other guides cover adjacent choices. OutsourcingVN is operated by Netbase JSC and is Netbase's own outsourcing-services platform.

Related services and solutions