Unstructured Document → Parquet
Normalize document text into Parquet-ready JSON rows.
Inputs
Raw document text
Entry: #77, 2024-01-02, service fee, $19.50
Outputs
JSON rows for Parquet
{ ...structured output... }
How to run this template
1
Add inputs in column A.
2
Run the local template on column B.
3
Review the output and export when ready.
Common use cases
Data lake ingest
Normalize document text into Parquet-ready JSON rows.
Analytics pipelines
Normalize document text into Parquet-ready JSON rows.
Warehouse loads
Normalize document text into Parquet-ready JSON rows.
Template notes
Prompt: Return a JSON array of rows with fields: record_id, date, description, amount.
FAQ
Does this run locally?
Yes. Templates use local models that run in your browser.
Can I edit the prompt?
Yes. Open the column settings and adjust the prompt.
What output format is expected?
The template includes a recommended output format you can tweak.