Capture. Resolve.
Route. Deploy anywhere.
The CDI layer between your data sources and every tool that needs customer context. Collect behavioral events, resolve identities across touchpoints, route enriched data to your entire stack. Self-hosted. Zero data egress.
Your data stack stores events.
It doesn't know your customers.
Events arrive from your website, mobile app, backend systems, and data warehouse — but they sit in separate tables, attached to different IDs, with no tool connecting them into a coherent picture of the same person.
Analytics queries stale data
Sessions, events, and identifiers land in different tables with no reconciliation layer. Your dashboards lag and your segments are weeks behind reality.
AI agents hallucinate
The context your agents receive is incomplete — fragmented device sessions, missing purchase history, unresolved anonymous visitors. Bad context produces bad outputs.
Marketing personalizes fragments
Your team segments against incomplete profiles. The same customer appears in three different audiences, or not at all, because identity was never stitched.
Meiro Pipes is the CDI layer that closes this gap. It sits between your data sources and every downstream tool — collecting events, resolving identities across touchpoints, and routing clean, enriched data wherever your stack needs it.
The full data control plane
for engineering teams
Pipes handles collection, identity resolution, event routing, warehouse modeling, and operational visibility — as one backend, inside your infrastructure.
Collect events from every touchpoint
Event stream sources, Web SDK auto-tracking, and Cloud Apps from external APIs. Source transform functions normalize raw payloads into typed events.
Resolve identity across channels
Configurable Identifier Types with merge priority and max-identifier limits. 184ms median resolution latency. Overflow profiles prevent pathological super-identities.
Route to any destination
A Pipe connects a Source to an Event Destination through an optional JavaScript transform. Destinations are custom send() functions — write once, deliver anywhere.
Feed Meiro Audiences for enrichment
What Pipes collects, Meiro Audiences profiles immediately. Same identity graph, same data layer — no syncing between products. Add Audiences when your team is ready.
Built for engineering teams
who own their data stack
Data Collection
Three source types — Event stream, Web SDK, Cloud Apps. Event stream and Web SDK sources have a JavaScript transform function that normalizes raw payloads into typed, validated events before storage. Reverse ETL sources define events via SQL query.
- Event stream: HTTP POST collection endpoint
- mpt.js: auto-tracks 40+ event types
- Cloud Apps: scheduled API pull imports
- JSON Schema validation
Identity Resolution
Define Identifier Types with configurable maxIdentifiers and merge priority. Background workers resolve profiles at 184ms median latency. Overflow profiles prevent bad merges on shared devices.
- Configurable merge rules per identifier type
- Priority-based conflict resolution
- Event deduplication
- Overflow profile isolation
Event Routing
Pipes connect Sources to Event Destinations via optional JavaScript transforms. Filter, reshape, or enrich events. Destinations are custom send() functions. Routing is separate from profile activation — keeping your event stream clean.
- Per-Pipe JS transform functions
- Custom HTTPS delivery targets
- AES-256-GCM encrypted secrets
- Full delivery history per Pipe
Warehouse Modeling
Native Snowflake, BigQuery, Databricks, PostgreSQL, MySQL, and MSSQL connections. Warehouse Models map existing tables to Identifier Types — historical data participates in the same identity graph as real-time events. No ETL, no export, no copy.
- Snowflake · BigQuery · Databricks · PostgreSQL · MySQL · MSSQL
- Zero-copy warehouse integration
- DWH-backed attributes in Studio
- Model-based identity resolution
Every Pipes instance ships with Piper, an AI agent with full read access to your CDI configuration. Autonomous heartbeat checks, anomaly detection, Slack/Teams/Discord alerts, and proposed fixes — every edit requires human approval. Available via product UI, mpcli, and external AI agent skills.
Deploy where your data
needs to live
No proprietary cloud dependencies. No MTU-based pricing traps. Customer PII never leaves your jurisdiction in self-hosted and any-cloud configurations.
| Deployment model | What it means | Data sovereignty |
|---|---|---|
| Self-hostedRecommended | Your Kubernetes cluster on AWS, GCP, Azure, or on-premise. Your cloud, your keys. | Zero data egress to any Meiro infrastructure |
| Any cloud, your region | Meiro deploys and manages in the cloud provider and jurisdiction you specify. You own the data layer. | Zero data egress — your jurisdiction |
| Meiro Managed SaaS | Fully managed by Meiro for teams that need fast deployment without infrastructure control. | Meiro-managed region |
AI agents (Piper and external) query customer context inside your infrastructure — no data export, no third-party model training on your customer data.
Start with Pipes.
Add layers when you're ready.
Pipes is the core data engine of the Meiro platform. It runs alongside Meiro Audiences on a shared backend — same identity graph, same data layer, no syncing between products.
| Product | Role | Handles |
|---|---|---|
| Meiro PipesYou are here | Captures and resolves the context | Events · Identity graph · Routing · Warehouse models |
| Meiro Audiences | Profiles and governs the context | Unified profiles · Attributes · Audience segmentation |
| Meiro Engage | Activates the context | Email · SMS · Push · Paid media · Journey Canvas · Agent-triggered actions |
Why teams choose Pipes over
Segment and RudderStack
Segment costs more, does less, and locks your data on Twilio's cloud. RudderStack gives you a pipe. Pipes gives you the pipe plus persistent profiles and activation.
| Tool | Data control | Identity resolution | Activation | Pricing model |
|---|---|---|---|---|
| Segment (Twilio) | SaaS only — Twilio cloud | Limited | No native activation | MTU pricing |
| RudderStack | Self-hosted option | No persistent profiles | No native activation | Event volume |
| Hightouch | Warehouse-dependent | No standalone capture | Reverse ETL | Rows synced |
| Meiro Pipes | Self-hosted · zero egress | Deep · configurable | Events + profiles + activation | 60–80% lower at scale |
Used in production at banks,
retailers, and travel companies
Deployed on-premise in a Kubernetes-managed environment to meet Czech financial regulations. Zero data egress.
Identity stitching across 6 brands and domains. Better-targeted campaigns through unified customer profiles.
Ask Meiro Pipes anything
