The workflow at a glance
- Atlantis e-shop fields → XML export
- File validation → staging catalogue
- Completeness approval → publication
What we need before implementation
- Confirmed Atlantis version and enabled e-shop module
- Organised categories and manufacturer information
- Controlled file delivery and a defined XML schema
At a glance
Integrate the store with Atlantis ERP around a defined workflow and the installation's available interface. Published Atlantis documentation provides a specific starting point: organised catalogue content and XML export. Distinguish internal identifier, published identifier, title, categories, attributes, images, translations and price.
What the integration should accomplish
Published Atlantis documentation provides a specific starting point: organised catalogue content and XML export. Distinguish internal identifier, published identifier, title, categories, attributes, images, translations and price. Discovery defines which fields belong to ERP and which are maintained by the store content team. Before publishing a new file, validate completeness so that a failed export cannot erase the existing catalogue or replace correct prices.
Interface and product boundaries
The official manual describes e-shop item content including categories, prices, images, attributes, and translations, together with XML file generation. It describes enabling e-shop in company settings and configuring the relevant lookup tables. The public catalogue documentation is older and does not itself prove a modern REST API or bidirectional order entry. Those capabilities must be checked in the actual installation before agreeing project scope.
A practical hypothetical scenario
A furniture business uses a published identifier different from its warehouse identifier. XML enters staging first, where categories, images, weight and available translations are checked. A new Greek description must not erase the English description simply because a file omits it. Validate the selected published price and whether it includes tax before changing the catalogue. An incomplete file stays unpublished, with an operator report and the previous valid catalogue preserved.
Failures, retries and ownership
An important failure scenario occurs when an incomplete XML file is generated when an export is interrupted. Keep the external reference and confirmed destination ID before retrying. Separate access failure, business rejection and an unknown outcome. An operator must be able to inspect the relevant record and resolve its status without replaying completed steps. Agree which fields may appear in logs and who handles unresolved cases.
Acceptance checks before launch
Test an incomplete file, a translation missing an image, and the correct published price. For each case record the expected source data, destination record and customer-visible state. Include one denied-permission case and one interrupted response. Re-run a previously confirmed event and check that its business effect is unchanged. The acceptance record should identify the responsible team and the evidence required to close any discrepancy.
Scope, handover and continuing operation
The initial discovery should confirm: Confirmed Atlantis version and enabled e-shop module; Organised categories and manufacturer information; Controlled file delivery and a defined XML schema. Before launch, document the mapping for Atlantis e-shop fields → XML export → File validation → staging catalogue → Completeness approval → publication. Assign responsibility for access renewal, monitoring and correction. A change to product configuration or the agreed business rule should trigger targeted regression checks. Begin with this defined workflow; add another channel only after its ownership and acceptance cases are clear.
Acceptance checks before launch
- an incomplete file, a translation missing an image, and the correct published price
- Repeated event without unintended duplicate effects
- Readable rejection and recovery status
- Source/destination sample reconciliation
FROM READING TO ACTION
A clear plan for your platform.
Tell us where your platform struggles and what needs to improve. Our team can connect the diagnosis with code, performance and infrastructure work.
Discuss your platform