userarena1002/mike
Parallel Databricks-target app added alongside the upstream codebase
userarena1002's sole commit drops a self-contained databricks_app/ subtree - FastAPI + Vite React + Databricks notebook jobs + Unity Catalog bootstrap scripts - next to the untouched upstream frontend/ and backend/ directories. The model layer is pluggable (mock | openai | databricks) but all Databricks job paths currently return mocks. A migration inventory documents the planned cuts: auth, sharing, R2/S3, Cloudflare deployment, free-form chat, and doc generation are all removed in favor of a single-user, upload-first review shape.
What's in it
- Parallel Databricks app A second target app aimed at Databricks, built alongside the original rather than replacing it.
- Migration kept reversible The upstream app is deliberately preserved so the Databricks version can prove itself before anything is cut over.
- Migration documentation Notes on how the Databricks move is being approached, sitting alongside the POC itself.
Direction
infrastructureintegration
Activity
Threads of work (detailed view)
userarena1002 is rebuilding Mike to run on Databricks
A from-scratch parallel version of Mike, aimed at enterprise data infrastructure and a stripped-down, single-user review tool.
userarena1002 adds a parallel Databricks-target app alongside the upstream codebase
A 6,687-line commit lands a full parallel application under `databricks_app/` - FastAPI backend, Vite React frontend, Databricks notebook jobs, Unity Catalog bootstrap scripts, and a migration inventory document - while leaving the original `frontend/` and `backend/` trees intact. This is a reorientation toward a specific enterprise deployment shape, not a drop-in feature.