Workflow Steps
Workflow steps are the building blocks of PlaidCloud automation. Each step performs one operation (import a CSV, join two tables, send a notification) and steps are composed into workflows that run sequentially, in parallel, conditionally, or in loops.
Categories
Section titled “Categories”Data Movement
Section titled “Data Movement”- Import — pull data in (26 source types: CSV, Excel, Parquet, JSON, SQL, BigQuery, SAS, SPSS, Singer sources, and more)
- Export — push data out (12 destination types: CSV, HTML, XML, SQL, Excel, Google Sheets, table archive, and more)
Table Transformations
Section titled “Table Transformations”- Tables — joins, unions, filters, lookups, pivots, melts, in-place updates (19 steps)
- Common operations — shared data mapping and filtering patterns
Modeling
Section titled “Modeling”- Allocation — driver-based spreading, split allocations, rule-based tagging
- Dimensions — load, sort, clear, export, and manage hierarchies
- Optimization — linear and mixed-integer solver steps
Documents and Files
Section titled “Documents and Files”- Document — file operations, PDF manipulation, image processing, encoding conversion (20 steps)
Communication
Section titled “Communication”- Notifications — email, Slack, Teams, SMS, webhook, Twitter, log (9 steps)
- Reports — render single or batch PDF reports from RML templates
Workflow Control
Section titled “Workflow Control”- Workflow control — variables, loops, sub-workflows, Macros, error handling (13 steps)
- General — LLM step, pass, wait, user-defined transform, run remote Python