Mosaic: A Framework for Geospatial Analytics at Scale
summary
Databricks outlines a concrete stack for large-scale geospatial work: Mosaic as a Databricks Labs framework for distributed spatial joins and indexing, Delta Lake for unified batch/stream storage, and delta sharing to publish large datasets cloud‑agnostically. Ingests WKB/WKT/GeoJSON today with GeoParquet planned, uses an index‑first strategy (H3 and British National Grid support) to scale complex polygons, recommends Apache Sedona for spatial SQL, and exports preprocessed layers to CARTO for visualization.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 0:47Why do you focus on geospatial data
- 1:51Databricks vision
- 5:24Why are we here today
- 6:03Mosaic
- 8:53Questions
related talks
46mGeoSpatially enabling your Big Data Apache clusters with LocationTech
05mDatabricks Product Announcements
23mKeynote Analysis
21mIbis, DuckDB, and GeoParquet: Making geospatial analytics fast, simple, and Pythonic
23mAnalystANGLE
11mTowards a Digital Society: Investigating Integration of Distributed DBs and Distributed Processing
24mManaging Heterogeneous Spatial Data with Analysis Ready Cloud Optimized formats
22mSanjeev Mohan, SanjMo & Tony Baer, dbInsight | Databricks Data + AI Summit 2024
