- Market
- Saudi Arabia
- Written for
- CIO, CTO
- Decision
- Integrate Oracle E-Business Suite or Fusion with surrounding systems
Why the checkpoint exists
Why Oracle makes you load through a gate you cannot skip
We do not have an office in Saudi Arabia
We serve organisations there from our Riyadh office and our engineering hubs in Lahore and Karachi, travelling for the parts of an engagement that need to happen in person.
Oracle’s base tables are not designed to be written to by anything except Oracle. A financial or supply-chain transaction is not a single row; it is a set of related records with validations, defaults and derived values that the application layer maintains together. Insert into the base tables directly and you get a row that looks right and is subtly broken — a transaction missing its accounting, a receipt with no matching lines, a record the rest of the suite cannot process. This is why Oracle ships open interface tables and import programs: the interface table is where you put candidate data, and the import program is what turns candidates that pass validation into real transactions.
Fusion, being cloud, removes the temptation by removing the access: you cannot reach its database, and integration happens through published REST and SOAP services, file-based import and Oracle Integration Cloud. That is the same discipline enforced by architecture rather than by convention. The move from E-Business Suite to Fusion is therefore not only a functional upgrade but a change in how every integration is built, and an integration written against EBS internals is exactly the one that has to be rebuilt rather than carried across.
One load, in detail
How an interface reaches E-Business Suite or Fusion without corrupting it
A safe load is a short, bounded sequence with validation in the middle and a rejection path for anything that fails. The steps differ a little between EBS and Fusion, but the shape is the same: nothing becomes a real transaction until Oracle’s own program has passed it.
Stage the data outside the base tables
On EBS the candidate rows go into the open interface table for that entity; on Fusion they are prepared as a file-based import or a REST payload. Either way they sit outside the real transactions until validated. Candidate data that cannot yet affect a posting.
Run Oracle’s own import and validation
The standard import program checks every row against the same rules a user’s entry would face, applying defaults and derivations the base tables do not carry on their own. Only valid rows advancing toward a transaction.
Handle the rejects explicitly
Rows that fail are reported with a reason, not silently dropped, so a data problem is corrected and reloaded rather than discovered later as a gap. A load whose failures are visible and fixable.
Schedule around the batch windows
On EBS the import runs as a concurrent request, so it is timed not to collide with the payroll, close or planning jobs the business already depends on. An integration that does not slow the overnight run.
The path
From a source system to a posted Oracle transaction, one validated step at a time
The diagram shows why the source system never touches a base table. Each stage hands off to the next only what has passed, so the transaction that finally posts has cleared the same gate a user would.
01
Source system
A CRM, a field app or a bank feed produces records that need to become Oracle transactions.
- CRM
- Field
- Bank
02
Staging or REST payload
Records land in an open interface table on EBS, or as a file-based import or REST call on Fusion — never in the base tables.
- Interface table
- FBDI
- REST
03
Oracle validation
The standard import program applies Oracle’s own rules, defaults and derivations, and rejects what does not pass.
- Import program
04
Posted transaction
Only validated data becomes a real transaction, complete with the accounting and links the suite needs to process it.
- GL
- Sub-ledgers
Reject and report
Anything that fails validation returns with a reason for correction, rather than being forced in. A load that has no rejection path is one that has been told to skip the validation, which is the failure mode.
The base tables are Oracle’s to write; every integration meets Oracle at the interface, not beneath it.
Our role
What we build against Oracle and what your DBA and finance team retain
- 01
Loads through the interfaces, not the tables
We build every inbound flow through EBS open interfaces and import programs, or Fusion REST and file-based import, so Oracle’s validation applies to our data as it does to a keyed entry.
- 02
Extraction that respects the batch schedule
We time and shape extracts so reporting and downstream feeds do not collide with the concurrent-manager windows your close and planning jobs depend on.
- 03
Configuration and flexfields stay yours
How your flexfields, value sets and accounting rules are set is your Oracle team’s domain; we build integrations that populate them correctly rather than working around a configuration we do not own.
Oracle’s configuration, its accounting rules and its batch schedule are your team’s and your implementation partner’s to own; we build integrations that load and extract through the interfaces Oracle supports without corrupting the data or crowding the batch, and leave interfaces your team can maintain.
Common questions
What buyers ask
What is the difference between integrating Oracle EBS and Oracle Fusion?+
The principle is identical and the mechanics are different. E-Business Suite is typically on your own infrastructure, so you can reach its open interface tables and run its import programs directly, and the discipline is to use those rather than the base tables beneath them. Fusion is cloud, so there is no database to reach at all; integration happens through published REST and SOAP services, file-based import and Oracle Integration Cloud. Fusion enforces by architecture what EBS enforces by good practice. The practical consequence is that an integration built for EBS does not simply move to Fusion — the flows have to be rebuilt against the published services, which is a planned part of any EBS-to-Fusion programme rather than an afterthought.
Can you load data straight into Oracle’s tables to save time?+
On EBS it is technically possible and it is the mistake we exist to prevent; on Fusion it is not even possible, which is Oracle telling you the answer. A base-table insert skips the import program’s validation, so it produces transactions that look complete and are subtly wrong — missing accounting, broken links, records the rest of the suite cannot process. The time it saves at load is repaid many times over when a reconciliation fails or a period will not close because a batch of transactions was never properly formed. Loading through the interface tables and import programs is slightly more work and the only route that leaves the data trustworthy.
Our extracts slow down the overnight batch. How do you avoid that?+
By treating the concurrent-manager schedule as a constraint the integration has to fit, not an environment it can ignore. Heavy extracts that run at the same time as payroll, period close or planning compete for the same resources and stretch the window everyone depends on. We schedule extracts around those jobs, shape them to pull only what changed rather than re-reading everything, and where the read is genuinely heavy we move it off the transactional system onto a reporting layer or a replica so it does not touch the batch at all. The aim is that a reporting query is never the reason a close ran late.
In this cluster
Next step
Design the Oracle interfaces
A session that decides, per interface, whether it loads through EBS open interfaces or Fusion REST, and how extraction is scheduled to avoid the concurrent-manager windows the business relies on.

