Hi @dbernstein_tp ,

You can use commission of Databricks Apps + Lakebase. You build a lightweight Streamlit/Dash front end, back it with a Lakebase (managed Postgres) table synced from Unity Catalog, and let users edit rows through the app.

There's also excel plugin, but you want to stop using spreadsheet:)

https://docs.databricks.com/aws/en/integrations/excel-write-back

If my answer was helpful, please consider marking it as accepted solution

Hi Dave!
This is a very common challenge across enterprise organizations looking to eliminate "spreadsheet sprawl" and brittle manual aggregation processes.
When it comes to enabling business users to read and write back data directly to Lakehouse tables in a spreadsheet-like interface, here are the primary architectural approaches:

1. The Native Approach: Databricks Apps (Streamlit / Gradio)
If you are already on Databricks, Databricks Apps is arguably the cleanest and most cost-effective solution without adding third-party SaaS vendors.
How it works: You can build a lightweight interactive web app hosted natively inside your Databricks workspace using Streamlit. Streamlit’s native st.data_editor widget renders an editable, Excel-like grid directly in the browser.
Why it shines: Direct Delta Lake Write-Back: Edits can immediately trigger PySpark / Delta Lake updates (MERGE INTO or appends) straight to your governed Unity Catalog tables.
Unified Security & Governance: It respects your existing Databricks SSO/Entra ID, RBAC, and Unity Catalog permissions.
No Extra License Fees: You only pay for the underlying compute required to run the app.

2. Third-Party BI / Write-Back Platforms
If business stakeholders strictly demand an enterprise BI tool with native spreadsheet write-back features out of the box:
Sigma Computing: As you mentioned, Sigma is well-regarded for its spreadsheet-native interface and direct write-back capabilities. It works well if non-technical business users need to build those input forms themselves without engineering support.
Power Apps / Power Automate + Power BI: If your company is heavily in the Microsoft ecosystem, building a simple Power App visual embedded directly inside a Power BI report is another popular route for structured data entry back into Delta tables (via Databricks SQL Warehouse or REST APIs).

3. Open Source / Embedded Grid Components
For teams that want custom internal tools without building from scratch:
Tooljet / Appsmith / Retool: Low-code internal tool builders that can connect to Databricks via SQL endpoints or REST APIs to create editable table interfaces.

Recommendation:
Before committing to a new SaaS procurement cycle for third-party tools, I strongly recommend building a quick Proof-of-Concept (PoC) using Databricks Apps with Streamlit.
In less than 50 lines of Python code using st.data_editor, you can deliver a secured, spreadsheet-style UI that reads and writes back to Delta Lake seamlessly under Unity Catalog governance.
Hope this helps guide your decision!

Thanks, great information!