Product portfolio
Five engines, one operational model
Choosing a different database should not mean learning a different operations story. Provisioning, credentials, backups, monitoring, access control and billing behave the same way across every engine WebyneDB manages.
| Service | Engine | Primary use | Status |
|---|---|---|---|
| WebyneDB PostgreSQL | PostgreSQL | SaaS, ERP and enterprise applications | Available |
| WebyneDB MySQL | MySQL | Web, e-commerce and line-of-business applications | Available |
| WebyneDB MariaDB | MariaDB | MySQL-compatible workloads | Available |
| WebyneDB Valkey | Valkey | Cache, sessions and queues | Available |
| WebyneDB Vector | PostgreSQL + pgvector | AI, RAG and semantic search | Available |
| WebyneDB Enterprise HA | Selected engines | Mission-critical production | On the roadmap |
PostgreSQL
SaaS, ERP, enterprise apps
The reference engine for WebyneDB. Everything the platform does — provisioning, TLS, backups with verified restores, streaming replicas, metrics and AI diagnostics — is deepest here.
Well suited to
- SaaS and multi-tenant applications
- ERP and line-of-business systems
- Workloads that need strict correctness and rich SQL
- Anything that will later want read replicas
What you get
- Versions
- 17, 16, 15, 14
- Managed endpoint
- Yes
- TLS
- Enabled
- Verified backups
- Yes
- Read replicas
- Yes
- Point-in-time recovery
- Phase 2
- Named database users
- Yes
PostgreSQL (pgvector)
AI, RAG, semantic search
PostgreSQL with the pgvector extension enabled at provisioning time, so embeddings live next to your relational data and one query can filter and rank together.
Well suited to
- Retrieval-augmented generation over your own documents
- Semantic and similarity search
- Recommendation and deduplication features
- Teams who would rather not run a separate vector database
What you get
- Versions
- 17, 16, 15
- Managed endpoint
- Yes
- TLS
- Enabled
- Verified backups
- Yes
- Read replicas
- Yes
- Point-in-time recovery
- Phase 2
- Named database users
- Yes
MySQL
Web, e-commerce, applications
Managed MySQL with the same control-plane abstractions as PostgreSQL: plans, endpoints, credentials, backups, monitoring, maintenance, scaling and billing.
Well suited to
- Web applications and content platforms
- E-commerce systems
- Existing MySQL applications being moved off self-managed servers
What you get
- Versions
- 8.4, 8.0
- Managed endpoint
- Yes
- TLS
- Enabled
- Verified backups
- Yes
- Read replicas
- Phase 3
- Point-in-time recovery
- Not applicable
- Named database users
- Yes
MariaDB
MySQL-compatible workloads
MariaDB for MySQL-compatible workloads, managed identically to every other engine on the platform.
Well suited to
- MySQL-compatible applications
- Workloads standardising on MariaDB
- Migrations from older MySQL versions
What you get
- Versions
- 11.4, 10.11
- Managed endpoint
- Yes
- TLS
- Enabled
- Verified backups
- Yes
- Read replicas
- Phase 3
- Point-in-time recovery
- Not applicable
- Named database users
- Yes
Valkey (Redis)
Cache, sessions, queues
Valkey — the open-source Redis-compatible engine — for the fast, ephemeral layer in front of your database.
Well suited to
- Application and query caching
- Session storage
- Rate limiting and job queues
What you get
- Versions
- 8, 7.2
- Managed endpoint
- Yes
- TLS
- Password auth
- Verified backups
- Yes
- Read replicas
- Phase 3
- Point-in-time recovery
- Not applicable
- Named database users
- Single shared password
MySQL and MariaDB on the same control plane
MySQL and MariaDB use the same customer experience and control-plane abstractions as PostgreSQL — plans, endpoints, credentials, backups, monitoring, maintenance, scaling and billing. Engine-specific providers implement the actual lifecycle operations underneath.
- Standalone deployments are supported today.
- Engine version policy is explicit, and upgrades are tested before being offered.
- Replication and high availability follow once backup and restore workflows are production-tested — deliberately, in that order.
Not sure which engine fits?
Describe the workload and we will tell you honestly which one we would run it on.