you basically have a nested array inside `input.values`, where each element is itself an array, and the first element of that sub-array is an object containing `"email"`.

In Azure Data Factory (ADF) or Synapse Data Flows, you can flatten this cleanly without multiple copy activities.

Here’s a step-by-step approach using a Mapping Data Flow:

---

Source

• Point your Source dataset to the JSON file (or API output).

• In Projection, make sure the schema is imported so you can see `input.values`.

---

First Flatten

• Add a Flatten transformation.

• Unroll by: `input.values`

This will give you each inner array as a row.

---

Second Flatten

• Add another Flatten transformation.

• Unroll by: `input_values[0]` (the first element of the inner array — the object with `email`).

• Now you can directly access `email` as a column:

`input_values[0].email`

---

Select Only Email

• Add a Select transformation.

• Keep only the `email` column.

Sink

• Set your Sink dataset to CSV.

• Map `email` → `email` in the output.