PIPELINE MODES
Structured & Raw Pipelines
Structured path: normalization, ABI decoding, and SQL transforms with at-least-once delivery. Raw path: byte-level passthrough to Kafka with exactly-once delivery.DATA QUALITY
Normalized Data
Every entity type — logs, transactions, traces — mapped to well-defined columns with predictable types. Consistent schemas across all supported chains.DECODING
ABI Decoding
Decode raw event logs and transaction input against a provided ABI. Matched entries expand into named, typed columns. Unmatched entries can be skipped or routed to fallback tables.TRANSFORMS
SQL Transforms
Filter, project, reshape, join across tables, and aggregate with standard SQL — applied per table after decoding, before data reaches the destination.DESTINATIONS
Multiple Destinations
Current destination types include ClickHouse, Postgres, Kafka, Object Storage, AWS SQS, and Webhook.CHAIN SUPPORT
Protocol-Native Schemas
Purpose-built normalizers for EVM chains, Hyperliquid, and Solana deliver rich, structured data with a deep understanding of each ecosystem's specific needs — not raw byte arrays.