Open Source Data Movement Platform
Migrate, replicate, and sync databases in real-time. Zero per-row charges. Your data never leaves your network.
0K+
Rows per second
0.9%
Reliability
0x
Parallel workers
Connectors
21 production-ready connectors with full CDC support. 5 more on the roadmap.
SQL Databases
Full schema discovery, cursor-based incremental sync, and automatic primary key detection.
SQL Databases
Binlog-based CDC with automatic position tracking and GTID support for failover.
SQL Databases
Change Tracking CDC with snapshot isolation and full incremental cursor support.
SQL Databases
LogMiner-based CDC for real-time change capture with supplemental logging.
Data Warehouses
Bulk loading via internal stage with auto-schema mapping and merge upserts.
Don't see your database? Build a custom connector with the Go SDK.
Browse connector docs →How It Works
01
Run a lightweight agent inside your network. One Docker command. No inbound ports, no VPNs, no firewall changes.
02
Paste a connection string. Zwiron auto-detects the database type, discovers schemas, and validates the connection instantly.
03
Pick tables, configure parallelism, set transforms, and hit start. Monitor everything live with per-row progress tracking.
Real-Time Monitoring
Watch your migration in real-time with per-table progress, worker stats, and throughput metrics.
Throughput
125,000
rows/s
Total Rows
23.4M
processed
Workers
8
active
Elapsed
3m 24s
Developer Experience
Start moving data with a few lines of code. Native Go SDK, REST API for everything else.
package main
import (
"context"
"fmt"
zwiron "github.com/zwiron-io/sdk-go"
)
func main() {
client := zwiron.NewClient("zw_your_api_key")
job, _ := client.Jobs.Create(
context.Background(),
&zwiron.JobParams{
Source: "postgresql://src:5432/app",
Destination: "snowflake://wh/analytics",
Mode: zwiron.ModeCDC,
Tables: []string{"users", "orders"},
Workers: 8,
},
)
fmt.Println("Job", job.ID, "started")
}Capabilities
Everything you need to move data reliably at any scale.
Capture every database change via WAL and binlog streaming. Full refresh, incremental, and CDC with automatic cross-run position resume.
Distribute work across up to 32 parallel workers per job for maximum throughput on tables of any size.
Auto-discover tables, columns, primary keys, indexes, and row estimates from any connected source database.
Every agent authenticates with mutual TLS. ECDSA P-256 certificates are issued and rotated automatically.
Column rename, filter, cast, and hash transforms built in. Cron scheduling with enable/disable and manual trigger.
WebSocket-powered dashboards with per-row progress, worker stats, and throughput metrics. Webhook delivery with retry tracking.
Security
Zero-trust architecture. Every byte is processed inside your network by the Zwiron agent.
0.99%
Uptime SLA
Enterprise tier
<0ms
Agent Latency
P99 gRPC round-trip
0
Data Exposure
Runs entirely in your VPC
0-bit
Encryption
AES-GCM at rest & in transit
The agent processes everything inside your network. Zwiron never sees, stores, or proxies your data.
100K+ rows/sec on commodity hardware. Scale to 32 workers per job.
Agent connects outbound only. No firewall holes, no VPNs, no SSH tunnels.
We never charge per row. Move a billion rows on any plan. Flat, predictable pricing.
CDC via WAL/binlog, WebSocket monitoring, and per-row progress.
Build custom connectors with a clean Go SDK. Plugin architecture.
Agent, engine, and connector framework are MIT licensed. Inspect every line of code that touches your data.
Pricing
No per-row charges. No hidden fees. No surprises.
All plans include unlimited data volume. Compare plans in detail →