Overview

End-to-end guides for canonical Apache Pinot workloads.

The rest of the Pinot documentation covers individual features: schema design, ingestion modes, indexes, query syntax, and operational knobs. These playbooks stitch those features together into complete, production-ready patterns for the workloads adopters ask about most often.

Each playbook covers:

Section	What it answers
When to use this pattern	Is this the right fit for my workload?
Architecture sketch	Which Pinot components are involved and how data flows
Schema and table config	Concrete JSON you can adapt
Ingestion setup	Stream and/or batch job configuration
Indexing strategy	Which indexes to enable and why
Query patterns	Representative SQL and performance tips
Operational checklist	Monitoring, scaling, and common pitfalls

Playbooks

Playbook	Typical use case
Real-Time Product Analytics	Sub-second dashboards over Kafka event streams (page views, clicks, transactions)
CDC / Upsert Pipeline	Keep Pinot in sync with a transactional database via Debezium and upserts
Hybrid Real-Time + Offline	Combine low-latency streaming with high-quality batch backfills in a single logical table
Multi-Tenant Analytics	Serve many customers from one Pinot cluster with resource and data isolation
Text Search Analytics	Full-text search over logs, product catalogs, or support tickets alongside OLAP aggregations

How to use these playbooks

1. Pick the playbook closest to your workload.

2. Copy the configs and adjust field names, topic names, and resource sizes for your environment.

3. Cross-reference the feature docs linked inside each playbook for deep dives on any single capability.

4. Check the operational checklist before going to production.

These playbooks target Pinot 1.x and assume familiarity with Pinot's architecture and table types. If you are new to Pinot, start with the 10-Minute Quickstart first.