https://github.com/databricks-solutions/model-landscape
Databricks-native model observability for teams that want production ML monitoring without moving inference data out of their workspace.
https://github.com/databricks-solutions/model-landscape
mlflow observability
Last synced: 30 days ago
JSON representation
Databricks-native model observability for teams that want production ML monitoring without moving inference data out of their workspace.
- Host: GitHub
- URL: https://github.com/databricks-solutions/model-landscape
- Owner: databricks-solutions
- License: other
- Created: 2026-05-26T17:21:26.000Z (about 2 months ago)
- Default Branch: dev
- Last Pushed: 2026-06-06T21:05:06.000Z (about 1 month ago)
- Last Synced: 2026-06-06T21:14:01.059Z (about 1 month ago)
- Topics: mlflow, observability
- Language: Python
- Homepage: https://databricks-solutions.github.io/model-landscape/
- Size: 1.4 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE.md
- Security: SECURITY.md
- Notice: NOTICE.md
Awesome Lists containing this project
README
# Model Landscape
Databricks-native model observability. Two halves of one product:
- **mlflow-lens** — a small SDK that enriches MLflow runs with interactive
Plotly panels (classification, regression, model selection), training-time
drift, structured summaries, and workspace context.
- **The warehouse app** — a Databricks App that monitors any Unity Catalog
inference table for drift, performance degradation, data-quality
regressions, and incidents. Data never leaves the workspace.
## Docs
The full documentation site is the canonical source — install guide,
architecture, SDK reference, panel gallery, and the intro deck all live
there:
→ **https://databricks-solutions.github.io/model-landscape/**
Run it locally:
```bash
uv sync --extra docs
uv run --extra docs python docs/_build_gallery.py
uv run --extra docs mkdocs serve
```
## Quick deploy
```bash
git clone https://github.com/databricks-solutions/model-landscape.git
cd model-landscape
uv build --wheel --out-dir dist
uv build --wheel --out-dir dist mlflow-lens
databricks bundle deploy -t warehouse_only \
--var "sql_warehouse_id=" \
--var "refresh_node_type_id=" \
--var "control_plane_catalog=" \
--var "control_plane_schema=model_landscape_control_plane"
databricks apps start model-landscape
databricks apps deploy model-landscape \
--source-code-path /Workspace/Users//.bundle/model-landscape/warehouse_only/files
databricks bundle run tutorial_mlops
```
Full first-time-setup guide: [Get started](https://databricks-solutions.github.io/model-landscape/getting_started.html).
## Local development
```bash
uv sync --extra dev # install workspace + dev deps
uv run --extra dev pytest -q tests
uv run --package mlflow-lens --extra dev pytest -q mlflow-lens/tests
uv run --extra docs mkdocs build --strict
uv build --wheel --out-dir dist
uv build --wheel --out-dir dist mlflow-lens
uv run python -m model_landscape.app # run app locally
```
## License
See [LICENSE.md](LICENSE.md). Third-party notices in [NOTICE.txt](NOTICE.txt).