The read layer
your database is missing
ReflexDB syncs your production database in real time and generates a typed REST API from your schema — no backend code, no infrastructure to manage.
How it works
Three steps from database to REST API.
Connect
Point ReflexDB at your MySQL, MariaDB, or PostgreSQL database. SQL Server support is coming soon. Supports direct connections, SSH tunnels for private databases inside a VPC, and read replicas.
Sync
ReflexDB reads your schema and streams changes via CDC (binlog, logical replication, or change tables depending on your database). Sub-second propagation. Schema changes are detected at the next build — trigger one from the dashboard or API.
Query
A typed REST API is deployed for you, scoped to your schema. Filter on any column, traverse relations, aggregate — with an interactive API explorer included at no extra cost.
Everything you need, nothing you don't
Built on managed cloud infrastructure. No proprietary lock-in.
Real-time sync
Sub-second propagation via MySQL binlog or PostgreSQL logical replication. SQL Server CDC polling is coming soon. Handles reconnects, position tracking, and backpressure. Falls back to poll mode automatically if CDC is unavailable.
Auto-generated REST API
Type-safe endpoints generated from your schema at build time. Filter on any column, traverse FK relations, sort, paginate, and aggregate — zero middleware to write.
Multi-region deployment
Deploy instances to Europe (Ireland) or US East (Virginia). Set a default region per connection and override per instance to put your data close to your users.
Auto-scaling
Enterprise plans scale horizontally to multiple replicas behind a load balancer automatically. Free tier pauses after two days of inactivity; paid plans run 24/7.
Private database support
Connect databases inside a VPC via SSH tunnel. API-key authentication on every endpoint. Data stays in your chosen region and is held in memory only — nothing is persisted to disk.
Management API
Full management REST API with long-lived API keys and an OpenAPI spec. Automate instance provisioning and configuration in your CI/CD pipeline.
Interactive API docs
Interactive API explorer and OpenAPI 3 spec auto-generated from your deployed schema. Import into Postman or Insomnia, or explore from the browser — included on every instance.
Enterprise ready
Audit log export, 99.95%+ SLA, dedicated support, and annual contracts. DPA and compliance documentation included on all paid plans.
Built for read-heavy workloads
One engine, many use cases.
API caching
Serve read-heavy endpoints from a ReflexDB replica instead of hitting your production database on every request. In-memory storage delivers near-zero latency with no cache warming required.
Search indexes
Maintain a filtered, typed read layer for search workloads. Use config-driven where predicates to select rows and fields to pick columns — without touching your schema.
Analytics dashboards
Run heavy aggregation queries against a live in-memory snapshot without adding load to your production database. No query planning overhead, no index tuning.
Mobile backends
Generate typed REST endpoints from your database schema and call them directly from your app. No custom middleware, no GraphQL resolver boilerplate, no ORM to maintain.
Start free. Scale when you're ready.
One nano instance included at no cost. Upgrade to Pro from $99/month for production workloads with real-time CDC sync and no rate limiting.
Frequently asked questions
How is this different from a MySQL read replica?
A read replica is still a database — you still need to write queries, manage connections, handle authentication, and build an API layer on top. ReflexDB adds all of that: a REST API generated from your schema, API-key auth, filtering, FK relation traversal, and aggregations. You query it with HTTP, not SQL.
What databases does ReflexDB support?
MySQL 5.7+, MariaDB 10.3+, and PostgreSQL 13+ are supported on all plans. SQL Server 2016+ support is coming soon. All backends work with managed database services, self-hosted instances, and databases inside a private VPC via SSH tunnel.
How fresh is the data?
With real-time CDC, changes typically propagate in under a second under normal load. Poll mode syncs on a configurable interval — minimum 300 seconds on Free, scaling down by tier to 5 seconds on Enterprise. CDC mode propagates changes in near real-time regardless of this setting.
Where is my data stored?
Each ReflexDB instance runs in your chosen region (Europe and US East available at launch). Data is held entirely in memory — nothing is written to disk. When an instance is paused or terminated, the in-memory data is discarded and re-synced from your source database on next start.
Can I connect a database inside a private VPC?
Yes. ReflexDB supports SSH tunnel connections for databases that are not publicly accessible. Configure an SSH jump host and provide your private key — the tunnel is established automatically during provisioning.
What happens if my production database goes down?
ReflexDB continues to serve the last synced snapshot. Read traffic is unaffected. Once the source reconnects, the engine resumes syncing from where it left off — no data is lost from the replica.