Workflow · Production
Hackathon speedrun: build and deploy a RAG app in minutes with Vertex AI Studio and Vertex AI Search
The problem
Hackathon participants need to quickly build and deploy a grounded LLM Q&A application from a knowledge base without complex setups.
Workflow diagram · grounded in source
1
Upload docs to Cloud Storage
integration
“Google Cloud Storage (GCS): To store our source documents.”
2
Index with Vertex AI Search
integration
“Vertex AI Search (within AI Applications): To index our documents and make them searchable.”
3
Ground prompt in Vertex AI Studio
ai_action
“Vertex AI Studio: To create our prompt, connect it to the Vertex AI Search data store (grounding), and test.”
4
Deploy as web app via Cloud Run
output
“Cloud Run: To deploy our AI assistant as a web app with a single click.”
5
Answer questions with grounding citations
ai_action
“You should see the AI generate an answer, and importantly, cite the "Grounding Sources" it used from your data store! This is RAG in action.”
Reported outcome
A fully deployed, interactive RAG application was built from raw documents with minimal coding using Google Cloud's Vertex AI Studio and Vertex AI Search.
Reported metrics
Deployment speedin what feels like minutes
Coding effortMinimal coding, maximum automation
Reported stack
Google Cloud StorageVertex AI SearchVertex AI StudioCloud Rungemini-2.0-flash-001GitHub
Source
https://mlops.community/blog/hackathon-speedrun-build-and-deploy-a-rag-app-in-minutes-with-vertex-ai-studio-and-vertex-ai-search
Read source ↗Frequently asked questions
What did this team achieve with this AI workflow?
A fully deployed, interactive RAG application was built from raw documents with minimal coding using Google Cloud's Vertex AI Studio and Vertex AI Search.
What tools did this team use?
Google Cloud Storage, Vertex AI Search, Vertex AI Studio, Cloud Run, gemini-2.0-flash-001, GitHub.
What results were reported?
Deployment speed: in what feels like minutes; Coding effort: Minimal coding, maximum automation (source-reported, not independently verified).
How is this workflow AI workflow structured?
Upload docs to Cloud Storage → Index with Vertex AI Search → Ground prompt in Vertex AI Studio → Deploy as web app via Cloud Run → Answer questions with grounding citations.