TL;DR
Production features (activity feeds, dashboards, cross-contract filtering) usually outgrow ad-hoc RPC queries. This guide walks through indexing Kaia with Goldsky using two approaches: Subgraphs (GraphQL APIs) and Mirror (replicating raw data into your own stack).
Start indexing: https://blog.kaia.io/indexing-kaia-with-goldsky/
When to use what
Subgraphs (GraphQL APIs)
Use this when you want a queryable API for:
- event-centric UX (feeds, histories, dashboards)
- stable query patterns for frontends
- fast prototyping (including ABI → “instant” API-style workflows)
Mirror (real-time replication)
Use this when you want raw data in your infrastructure for:
- complex analytics (SQL joins with off-chain tables)
- high-volume processing (owns the data layer)
- internal tooling /BI/backend workers
Quick start checklist
- Pick a path: Subgraphs (aggregation + GraphQL) vs Mirror (raw data into your DB)
- Install Goldsky CLI + create an API key
- Deploy: push your subgraph config or define a replication pipeline