Quick Start
Get LakeOps connected to your data lake and running autonomous optimization in under 10 minutes. No agents to install, no data movement, and no pipeline changes required.
1. Connect your catalog
LakeOps connects directly to your existing Iceberg catalogs. Supported catalog types include AWS Glue, Hive Metastore, REST catalogs, Polaris, Gravitino, Nessie, Lakekeeper, and S3 Tables.
Example — catalog setup
Glue + S3) and enter your connection details.You can connect multiple catalogs across different regions and cloud environments.
2. AI analyzes & simulates
Once connected, LakeOps models table health from metadata, query patterns, and cost signals. It identifies tables that need compaction, snapshot cleanup, manifest rewrites, or orphan file removal — and categorizes them as Critical, Warning, or Healthy.
Example — dashboard after connection
3. Enable automated optimization
Every optimization operation in LakeOps (compaction, snapshots, orphan cleanup, manifest rewrites) can run in one of two modes:
Auto (autopilot)
Toggle the Auto switch to on. LakeOps will run the operation automatically on the configured cron schedule. No manual intervention required.
Manual (on-demand)
Toggle the Manual switch to on. You control when the operation runs by clicking Save to persist settings, then triggering execution yourself.
You can find these toggles in the Explore > Optimization tab for any table. You can also apply rules at scale via organization-wide policies.
4. Monitor with full visibility
The Dashboard shows real-time optimization activity across your entire lake. Track total operations, query speed improvements, cost savings, and data optimized.
Example — dashboard metrics
Key guarantees
- ✓No vendor lock-in — LakeOps works with your existing catalogs, engines, and storage. Nothing proprietary is added to your data.
- ✓No code or infrastructure changes — connect via metadata; no agents, sidecars, or pipeline rewrites.
- ✓No data movement — LakeOps reads metadata and writes optimized files back to your storage (S3, GCS, ADLS). Your data never leaves your account.