[FAQ Bot] NEW: Does it matter which serverless database I pick for the bonus deployment

[github-actions]

✨ FAQ NEW

Course: llm-zoomcamp
Section: module-5 (The topic is about monitoring/feedback flow behavior and persistence, which aligns best with Module 5’s Monitoring content (user feedback handling), rather than general course logistics or RAG/orchestration modules.)
Related Issue: #344

Question

Does it matter which serverless database I pick for the bonus deployment points when user feedback arrives after the interaction is stored?

Decision Rationale

This specific question about delayed user feedback (INSERT followed by a later UPDATE) and whether the choice of serverless database matters is not covered by any existing FAQ in the provided SEARCH_RESULTS.

Placement Details

• Section ID: module-5
• Sort Order: End of section
• Filename Slug: serverless-db-delayed-feedback-insert-update

---

🤖 Generated by FAQ Bot

Closes #344

[alexeygrigorev]

Closing: this isn't in scope for the course FAQ.

Choosing between serverless database vendors for a capstone deployment is general cloud-architecture advice rather than LLM Zoomcamp material — neither BigQuery nor Neon appears in the course, and no evaluation criterion depends on the database you pick (the bonus is "Deployment to the cloud (2 points)" for deploying at all).

For the record, the technical claim was also out of date: BigQuery's DML restriction on recently written rows is 30 minutes and applies to tabledata.insertAll and the Storage Write API (REST); rows written over the Storage Write API (gRPC) or with a plain SQL INSERT can be updated immediately. The ~90-minute figure now only covers availability for copy operations. See https://docs.cloud.google.com/bigquery/docs/data-manipulation-language

Closing #344 alongside this.