Introduction
DBConvert Streams combines data exploration, SQL workflows, and stream execution in one product.
It has two layers:
- Explorer / SQL — browse schemas, preview data, run queries, generate diagrams.
- Streams — run one-time Load migrations or ongoing CDC replication.
Exploring data is always available. Running a stream is where trial limits and license entitlements apply.
An AI assistant works across both layers — ask about your workspace in plain language, inside the app or from your own AI client. See AI assistance.
Capability matrix
| Connector | Source | Target | Load | CDC | Explorer | SQL |
|---|---|---|---|---|---|---|
| MySQL | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| PostgreSQL | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Snowflake | ✗ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Local files | ✓ | ✓ | ✓ | ✗ | ✓ | ✓ |
| S3 storage | ✓ | ✓ | ✓ | ✗ | ✓ | ✓ |
| Federated queries | ✓ | ✗ | ✓ | ✗ | ✓ | ✓ |
- MySQL CDC uses binlog-based replication.
- PostgreSQL CDC uses logical replication.
- Local files supports CSV, JSON, JSONL, and Parquet as sources. Target formats are CSV, JSONL, and Parquet.
- Snowflake is a target only. Load stages Parquet and runs
COPY INTO; CDC streams changes through Snowpipe Streaming into a change table, which a Snowflake task merges into an ordinary current-state table. CDC requires key-pair authentication, a primary key on each table, andEXECUTE TASKon the account. - S3 storage covers S3-compatible APIs; does not imply parity with every object-store provider.
- Federated queries support multi-source SQL and query-driven Load workflows only, not continuous CDC.
Data exploration and SQL
Browse and query any connected source without running a stream.
| Scenario | Example | Setup |
|---|---|---|
| Schema inspection | View tables, columns, keys, indexes | Data Explorer |
| Data preview | Browse rows in a read-only view | Tables and Views |
| SQL queries | Run queries against databases or files | SQL Console |
| Cross-source queries | Join data from multiple connections | Federated Queries |
| Structure diagrams | Generate visual schema diagrams | Table Structure |
One-time conversion
Migrate or copy data between databases or file formats in a single run.
| Scenario | Example | Setup |
|---|---|---|
| Database migration | MySQL to PostgreSQL | MySQL Load, PostgreSQL Load |
| Schema conversion | Map MySQL types to PostgreSQL | Schema Conversion |
| File to database | Load CSV/Parquet into MySQL or PostgreSQL | Local Files, Stream Config |
| Database to file | Export tables to CSV, JSONL, or Parquet | Local Files, S3 Storage |
| Custom query transfer | Migrate specific rows with a query | Custom Query |
See Load Mode for how Load works.
CDC replication
Keep targets in sync continuously by capturing changes from MySQL or PostgreSQL as they happen.
| Scenario | Example | Setup |
|---|---|---|
| Database to database | MySQL to PostgreSQL | MySQL CDC, PostgreSQL Target |
| Database to Snowflake | MySQL or PostgreSQL to Snowflake | Snowflake CDC |
CDC sources are MySQL (binlog) and PostgreSQL (logical replication). CDC targets are databases and Snowflake; files and S3 take Load only. See What is CDC for background.
AI assistance
Ask about your databases, files, and streams in plain language. Every option is read-only — it inspects your live workspace (the same connections, schemas, and stream status you see in Data Explorer) and never changes DBConvert data or configuration.
| Scenario | Example | Setup |
|---|---|---|
| Chat inside the app | Ask why a stream stopped or what's in a database | AI Chat |
| Generate SQL | Describe a query and get SQL in the SQL Console | Generate SQL |
| Fix a failed query | Attach the failed SQL and error, review the suggested replacement | Fix failed SQL |
| Connect your own AI client | Use Claude Desktop, Cursor, or Copilot against your workspace | AI Clients (MCP) |
AI Chat runs your own agent CLI (Claude Code, Codex, GitHub Copilot CLI, or OpenCode) inside the desktop app; AI Clients (MCP) connect an external client over stdio or HTTP. Both read through the workspace that already powers Data Explorer, so there's no second copy of your schema to keep in sync.
Cloud and managed databases
All the workflows above work with hosted databases. Use these guides for provider-specific connection setup.
| Provider | Guide |
|---|---|
| Amazon RDS | RDS Setup |
| AWS Aurora MySQL | Aurora MySQL |
| AWS Aurora PostgreSQL | Aurora PostgreSQL |
| Google Cloud SQL | Cloud SQL |
| Azure Database | Azure Setup |
| DigitalOcean | DO Database |
| Neon | Neon Setup |
| Databricks Lakebase | Lakebase Setup |
| Supabase | Supabase Setup |
| YugabyteDB | YugabyteDB Setup |
Where to start
| Goal | Page |
|---|---|
| Install and run your first stream | Quick Start |
| Connect to a database or file source | Connection Management |
| Browse schemas and query data | Data Explorer |
| Ask AI about your workspace | AI Chat |
| Connect an external AI client | AI Clients (MCP) |
| Configure and run streams | Stream Configuration Guide |
| Automate via REST API | API Reference |
| Navigate the dashboard | Dashboard and UI Guide |