ELT product
Skippr Cloud ELT is the extract, load, and model product on Skippr Cloud — a sibling of tables, gateway, and functions. The runner discovers sources, syncs bronze data into your warehouse, and produces silver/gold dbt you can inspect and own.
Status: Production. Start with Install.
What it does
| Capability | What you get | Next |
|---|---|---|
| Data integration | Connect databases, files, and APIs; ingest with exactly-once delivery | Sources |
| CDC | Capture inserts, updates, and deletes with order-token guards | CDC overview |
| Schema discovery | Deterministic inference and backward-compatible evolution | How it works |
| Data cleansing | Schema evolution for drift; dead letters only for unparseable records | Operations |
| Data modeling | Dimensions, facts, and metrics as validated dbt | skippr model |
| AI-ready data | Typed Parquet and warehouse tables for downstream ML | Vision |
| Warehouses | Land in the warehouse you already use | Destinations |
| Features | Exactly-once, crash recovery, privacy, single binary | Core concepts |
How a first pipeline runs
- Install the runner — Install.
- Discover source metadata —
skippr discover. - Sync bronze data —
skippr sync. - Model reviewable dbt —
skippr model.
Row-level source data moves from the machine running skippr to your destination. Authentication and control-plane services are Cloud-backed. Cloud-backed usage bills as vCPU time, memory time, bytes stored, and network bytes — the same four meters as other Cloud capabilities. There is no monthly seat, MAR pack, or named usage-threshold SKU. See How it works and ELT features.
Related
- ELT home
- Cloud ELT APIs (managed pipeline APIs; Preview stub, not the runner)
- Compare ELT