CLI vs UI: When Each Makes Sense in Automation Tools

Should your automation tool have a CLI or a UI?

The answer: both. But build them separately.

When CLI Makes Sense

Use a CLI when:

• The tool runs on a schedule (cron jobs, cloud functions).
• The user is technical (developers, DevOps).
• You need scriptability (chaining commands together).

Example:

python process.py --file invoice.pdf --output results.xlsx

This is perfect for automation pipelines.

When UI Makes Sense

Use a UI when:

• The user is non-technical (analysts, operations).
• The task requires visual feedback (preview, progress bars).
• You need interactivity (upload files, adjust settings live).

Example: A Streamlit interface where users drag-and-drop PDFs and see results instantly.

The Best Approach: Separate Layers

Don't mix CLI logic with UI logic.

Your core processing function should look like this:

def process_file(file_path: str, output_format: str):
    # ... processing logic
    return result

Then build two interfaces on top:

CLI layer:

if __name__ == "__main__":
    import sys
    result = process_file(sys.argv[1], "xlsx")

UI layer (Streamlit):

uploaded_file = st.file_uploader("Upload PDF")
if uploaded_file:
    result = process_file(uploaded_file, "xlsx")
    st.write(result)

Same logic. Two interfaces.

My Default Pattern

I always build the CLI first.

Why? Because a CLI forces you to separate logic from UI.

Once the CLI works, adding a UI is trivial.

Conclusion

CLI for automation. UI for humans.

Build both. Keep them separate.