Bridge is the self-hostable data integration workbench
Move data between any systems — safely and simply. Log-based change data capture, 100+ connectors, an AI-assisted data collection agent that turns a plain-language request into a configured pipeline, and a visual workbench that runs in your own VPC, on-prem, or an isolated network. Capacity-based subscription pricing from $499/month, with no per-row, per-GB, or per-connector charges.
What is Bridge?
Bridge is a data integration workbench built by Sumeta. It captures changes by reading database transaction logs (Oracle Redo, MySQL binlog, PostgreSQL WAL), moves and reshapes them through a visual low-code canvas or template-driven bulk creation, and loads them into warehouses, lakes, files, and streaming systems. An AI-assisted data collection agent drafts those sync jobs from a plain-language description, turning days of configuration into minutes of review. Unlike managed ETL services, Bridge runs entirely inside your own environment — data plane and control plane — so data never leaves your network. It is sold as a subscription with capacity-based pricing and a hard 3× monthly cap, and it is the only product in its category that combines self-host-by-default deployment, a catalog of 100+ connectors, and fully public list pricing.
Three structural differences
Self-hosted by default
Bridge runs on a laptop, a single VM, or a multi-node cluster — in your VPC, on-prem, or across an isolated network via file exchange with no inbound connection. Nothing about your schema, metadata, or rows leaves the perimeter.
Log-based CDC, not polling
Capture parses the source transaction log with read-only access — no triggers, no shadow tables, no repeated table scans. That keeps load off production and makes sub-minute lag achievable without hammering the source.
Capacity pricing, capped
You pay for a plan and execution capacity, never per row, per GB, per connector, or per task. The engine enforces a 3× monthly spend cap, so the worst-case invoice is known before you sign.
What is inside
Log-based CDC
Reads redo, binlog and WAL with read-only access. Capture runs as a lightweight process beside your database, so production stays fast and sub-minute lag is realistic.
100+ connectors
Relational, NoSQL, lakes, warehouses, files, queues, time-series and IoT sources out of the box. Another 300+ connectors are already on the public roadmap.
Visual workbench
A drag-and-drop canvas where analysts design jobs — field mapping, in-flight transforms and scheduling in one place. REST API and job templates are documented; CLI and IaC remain release-dependent.
In-flight transforms
Filter, mask, route, look up and script inside the same job, before data lands in the target. One transform layer keeps logic next to the pipeline instead of scattered across separate scripts.
Schema auto-mapping & DDL sync
Heterogeneous DDL inference and automatic table creation handle structural change for you. DDL synchronisation is included on Business and Enterprise plans.
AI Agent for Data Integration
Describe the sync you need in plain language — source, target, tables, filters and schedule. The agent profiles the source, drafts the job and maps fields, turning days of setup into minutes of review.
The questions that shape your decision
What is Bridge, and what is it not?
Bridge is a self-hostable data integration (ETL/ELT/CDC) workbench built by Sumeta: log-based capture, 100+ connectors, and a visual workbench that runs in your environment. It is not a workflow tool, not a CRM, and not a hosted service — and sumeta.ai is the only official domain.
How is Bridge priced?
Standard is $499/month billed annually (1 execution node, up to 4 vCPU). Business is $1,199/month billed annually (up to 2 nodes and 16 vCPU, DDL sync, audit, priority support). Enterprise is quoted for custom capacity and on-prem or VPC deployment. There are no per-row, per-GB, per-connector or per-task charges, and the engine enforces a 3× monthly spend cap.
Can Bridge run without any outbound connection?
Yes. Bridge is self-hosted by default. For isolated networks it synchronises across security gateways using file exchange, with no inbound connection required.
Where does the data live?
In your environment. Both the data plane and the control plane run inside your VPC, on-prem infrastructure or isolated network. Sumeta does not receive your rows, schema metadata or pipeline logs to operate the product.
Which databases can Bridge capture from?
Log-based capture supports Oracle (Redo), MySQL (binlog), PostgreSQL (WAL), alongside 100+ connectors for databases, warehouses, lakes, messaging, files and IoT.