Skip to content

Add Document Analyzer App to ai_parse_document bundle#146

Open
garciatorres wants to merge 1 commit intodatabricks:mainfrom
garciatorres:add-document-analyzer-app
Open

Add Document Analyzer App to ai_parse_document bundle#146
garciatorres wants to merge 1 commit intodatabricks:mainfrom
garciatorres:add-document-analyzer-app

Conversation

@garciatorres
Copy link

Summary

  • Adds an interactive Document Analyzer App (Streamlit) to the job_with_ai_parse_document bundle
  • Users can upload a PDF and get instant analysis: document parsing (ai_parse_document), classification (ai_classify), and structured extraction (ai_query with Claude)
  • Uses the Databricks SDK Statement Execution API (REST) instead of databricks-sql-connector for reliable connectivity inside the Databricks App container
  • Adds warehouse_id bundle variable and app resource definition

Files added/changed

  • src/app/app.py — Streamlit application
  • src/app/app.yaml — App runtime config (command, env vars)
  • src/app/requirements.txt — Python dependencies
  • resources/document_analyzer_app.app.yml — DAB app resource
  • databricks.yml — Added warehouse_id variable
  • README.md — Updated with app documentation and setup instructions

Test plan

  • Deployed to Azure Databricks workspace and verified end-to-end
  • Tested PDF upload, parsing, classification, and summarization
  • Verified service principal permissions (catalog, schema, volume grants)
  • Confirmed SQL warehouse connectivity via Statement Execution API

This pull request was AI-assisted by Isaac.

Add an interactive Streamlit-based Databricks App that lets users upload
PDFs and analyze them on demand using ai_parse_document, ai_classify,
and ai_query. Uses the Databricks SDK Statement Execution API for
reliable SQL connectivity inside the app container.

Co-authored-by: Isaac
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants