> ## Documentation Index > Fetch the complete documentation index at: https://docs.planasonix.com/llms.txt > Use this file to discover all available pages before exploring further. # Nodes overview > Explore the complete library of pipeline nodes for sourcing, transforming, and delivering data. **Nodes** are the building blocks of every Planasonix pipeline. Each node performs one job: read data, reshape it, check quality, or write it somewhere else. You connect nodes with edges so outputs flow into inputs in a clear, reviewable graph. The library is grouped into **sources**, **row transforms**, **column transforms**, **joins and unions**, **aggregation**, **parsers and builders**, **control flow**, **data quality**, **advanced** code and engine integrations, and **destinations**. Browse by category below—each page lists node types, configuration fields you set in the inspector, and patterns that work well in production. ## Node categories Ingest from connections, tables, CDC, Iceberg, webhooks, and iterators. Filter, sort, sample, deduplicate, explode arrays, and Z-order sort rows. Dates, windows, schema mapping, type inference, and guided data prep. Combine datasets with joins, fuzzy matching, and unions. Summarize, pivot, and unpivot for analytics-ready shapes. Parse CSV, JSON, and XML; flatten nested structures; build payloads for downstream systems. Branching, errors, loops, splits, and cross-pipeline triggers. Validate, profile, detect PII, detect anomalies, generate test rows, and notify owners. SQL, scripts, notebooks, UDFs, embeddings, streaming, ML, and geospatial operations. Write to warehouses, lakes, cloud storage, and webhooks. ## How to choose a node If data is not yet tabular, use a **parser** or **source** that emits rows. If it is already relational, jump to transforms or joins. Built-in nodes carry clearer metadata for lineage and optimization. Use **Custom SQL** or **Python** when you need logic that no single node expresses cleanly. Add **Error Handler** or **Notification** nodes on critical paths so operational teams learn about issues before consumers do. ## Pipeline authoring resources Learn how to wire nodes, preview data, and debug runs. Parameterize node settings for reuse across environments.