Couchbase · NoSQL database
NoSQL database: what it is, the four types, and when an enterprise should choose one
A NoSQL database stores data without the fixed tables of a relational database, and most are built to spread data and load across a cluster of servers. That suits some enterprise workloads well and others not at all. This page explains the four types, when Saudi organisations choose one, and how Interkey assesses, builds and operates Couchbase, which serves key-value lookups and JSON documents queried with SQL++ in one platform.
Orientation
At a glance
The page in one table, before the detail.
- What NoSQL means
- Databases that store data without fixed relational tables: document, key-value, wide-column and graph
- Usual reasons to choose one
- Records that change shape, heavy operational load, growth across servers
- Usual reasons not to
- A relational system that serves its load, core multi-entity transactions, heavy reporting
- The platform Interkey implements
- Couchbase: key-value and JSON documents, SQL++, search, eventing and analytics services
- Where it can run
- Couchbase Capella in a cloud region, including Dammam, or self-managed Couchbase Server
- How it is bought
- A workload assessment first, then a quotation per programme
In short
- What it is: a NoSQL database keeps data as documents, key-value pairs, wide columns or graphs instead of fixed tables.
- Why enterprises use one: records that change shape, and load that one server cannot carry.
- Why some should not: if a relational system serves its load, or the core logic is relational and transactional, staying is usually right.
- Where Couchbase fits: one platform for key-value and document workloads, queried with SQL++.
What a NoSQL database is, in plain words
A relational database stores data in tables with fixed columns and joins them at query time. It is the right design for a great deal of business software.
"NoSQL" is a broad name for databases that store data another way. There is no single NoSQL design, but most share three traits:
- No fixed schema required. A record can carry new fields without a table change.
- Data stored the way it is read. Instead of joining several tables, the data a request needs is often kept together in one record.
- Built to scale out. Data and load are spread across a cluster of servers, and the cluster grows by adding nodes.
The trade-off differs by product: some offer SQL-based query languages and ACID transactions, others neither. Test the specific product against the queries and transactions your workload really runs.
NoSQL database examples: the four types
Each type answers a different question about the data. The first two map most directly onto the records business applications already handle.
| Type | What it stores | Typical enterprise examples | Watch for |
|---|---|---|---|
| Document | Self-describing records, usually JSON, read and written as a whole | Customer profiles, product catalogues, orders, case files, content | Modelling: copying tables one-to-one into documents usually performs worse |
| Key-value | A value looked up by a unique key | Sessions, caching, user preferences, fast lookups by a known identifier | Queries other than by key need another mechanism |
| Wide-column | Rows with many, sparsely filled columns | Very large volumes of writes and time-ordered data | Queries must be designed around the partition layout in advance |
| Graph | Items and the relationships between them | Networks, fraud rings, recommendations | A narrow tool: strong on connections, rarely the system of record |
Our SQL or NoSQL decision guide lists what to measure in a workload before choosing between relational and these four types.
NoSQL database in Saudi Arabia: when enterprises choose one
In the Kingdom, the decision rarely starts with a technology preference. It starts with a system that has outgrown its design: a relational database built for an internal application, now serving a public portal, a mobile app and an analytics team at once. A NoSQL database for enterprise use usually earns its place in four situations:
- National-scale portals and apps. Peaks around a launch, a deadline or a season that one database server cannot carry.
- Records that keep changing shape. Fields that change every release, or differ by type or source.
- Profiles, sessions and catalogues. Data fetched by key very often that must also be stored safely.
- Field and mobile work. Applications that must keep working with a weak or absent connection, then synchronise.
Read more
Two Saudi questions sit on top of the technical ones. The first is where the data may live: many organisations must keep certain data inside the Kingdom under the National Cybersecurity Authority's controls (nca.gov.sa) and the Personal Data Protection Law overseen by SDAIA (sdaia.gov.sa), and that rule decides the deployment model before any feature does. The second is who will run a distributed database for the next five years. A cluster without an owner is a future outage.
And sometimes the answer is to stay. If the relational system serves its load, or the core logic moves money or stock together, a move buys risk with little return. The data platform practice covers that wider decision.
Couchbase: one NoSQL platform for key-value and documents, with SQL++
Couchbase is a distributed NoSQL database. Interkey is a Couchbase technology partner and sells Couchbase Capella. According to Couchbase's documentation:
- Key-value and documents in one store. Each item is a JSON document or a binary value, identified by a key unique within its collection, fetched by key or queried.
- SQL++ for queries. JSON documents are queried with SQL++, a document-oriented query language.
- Transactions. ACID transactions are available through SQL++.
- Separate services. Data, query, index, search, analytics, eventing and backup run as services that can be deployed and scaled independently.
- Memory and disk. Data is held in memory and on disk, so one platform can serve hot reads and keep the data safe; see our distributed cache page.
- Replication between sites. Data can be replicated selectively between data centres.
- Mobile and edge. Couchbase Lite runs on devices and synchronises through App Services or Sync Gateway; see offline sync.
Read more
Capella, Couchbase's managed service, also lists full-text search and vector search indexes. That is why Couchbase is described as a multi-model option: several access patterns on one cluster.
Where it does not fit. Couchbase is not a graph database or a wide-column store, and a reporting warehouse belongs on an analytical platform. The Couchbase page has the full fit table.
Capella or self-managed: where your NoSQL database runs
Couchbase comes in two operating models, and in Saudi Arabia the choice is usually a residency decision before it is a convenience one.
- Couchbase Capella is the vendor's managed service on AWS, Google Cloud or Microsoft Azure. As of 2 November 2026, its published regions include Google Cloud's Dammam region (me-central2).
- Self-managed Couchbase Server runs on infrastructure you control, such as your own data centre in the Kingdom, and someone must size, patch and upgrade it for years.
Read more
A region inside the Kingdom settles where the cluster's data sits, not where backups, replicas or logs go; check those with your compliance owner. Our Capella or self-managed page compares the two.
Support comes from Riyadh, from Interkey engineers who work in Arabic and English, from the first workload review to operation after go-live.
RelatedCapella or self-managed page
How Interkey delivers a NoSQL database project
-
Workload assessment.
Access patterns, transaction scope, structural change, peak load and the residency rule. Its first legitimate outcome is "stay where you are".
-
Data modelling.
Documents shaped by how the application reads them, with keys and indexes for the real queries.
-
Platform build.
Capella in the agreed region, or Couchbase Server on your infrastructure, sized against measured load.
-
Migration while the old system runs.
Data synchronised continuously from the current database, usually with iblync, Interkey's own product; reads move service by service, then writes.
-
Rehearsed cutover.
Agreed pass conditions, a rollback that has been exercised, and a named person empowered to call it.
-
Operate or hand over.
Failover drills, upgrades and tuning by Interkey under a written managed-services scope, or a documented handover to your team.
What drives the cost of a NoSQL database
We do not publish a price or a duration; both depend on the workload. The licence follows the operating model:
- Capella is billed per node per hour, scaled by each node's vCPU and RAM; backups and data transfer are billed separately.
- Self-managed Couchbase Server is a subscription priced by the number of nodes in production.
The engineering cost is driven by other things:
Read more
- how many services read and write the data, which sets the size of a migration more than data volume does;
- how far the current model is from a document model;
- the transactions and consistency each business flow needs;
- how long the parallel run must last before writes move;
- who operates the cluster after go-live.
The assessment is a fixed piece of work, the migration is priced by the workloads in scope, and vendor licensing is shown separately.
Questions buyers ask
What is the best NoSQL database for an enterprise in Saudi Arabia?
The one that fits the workload and the hosting rule. Measure access patterns, transactions and peak load first. Couchbase is the platform Interkey implements, and the assessment can still recommend staying relational.
What are common NoSQL database examples in business systems?
Customer profiles and catalogues as documents, sessions as key-value pairs, high-volume time-ordered writes in wide-column stores, and fraud analysis in graph databases.
Can our NoSQL database stay inside the Kingdom?
Yes. Self-managed Couchbase Server can run in your own data centre in the Kingdom, and as of 2 November 2026 Capella's published regions include Google Cloud's Dammam region. Backups, replication and logs need the same check.
How is Couchbase licensed, and can we buy it through Interkey?
Capella is billed per node per hour, and Interkey sells it, so one quotation can cover Capella and our work. Self-managed Couchbase Server is a subscription by production node, quoted alongside our work.
Can our developers query Couchbase with SQL?
They use SQL++, a document-oriented query language for JSON based on SQL, so the starting point is familiar. Queries written for a relational schema still need review, because the data model changes.
How long does it take to put a workload on a NoSQL database?
It depends on how many services touch the data, how far the model is from documents and how long the parallel run lasts. The duration is estimated from the plan the assessment produces.
Next
Related pages
Next step
Check whether NoSQL fits your workload
Tell us what the system does, the database it runs on now, where it struggles and where the data must be hosted. We will reply with whether a NoSQL database fits.
Or directly
+966-11-2180999 [email protected]Tawuniya Towers, North Tower, 7th Floor, King Fahad Highway, Olaya, P.O. Box 56835, Riyadh 11564, Saudi Arabia