Observe Lake usage.
Look for repeated scans, joins, and aggregations.
Schedule jobs on FastBI Lake. AI analyzes Lake usage and recommends intermediate tables so repeated reports can scan less data and return answers sooner.
Several reports group raw orders by day and customer. AI identifies that repeated work in Lake usage.
lake.orders → group by day + customer01Find repeated work
When reports repeatedly scan the same raw data, AI can identify shared calculations and recommend an intermediate table worth building.
Look for repeated scans, joins, and aggregations.
Capture useful intermediate results in a reusable Lake table.
Let recurring reports query the prepared data.
02Keep tables fresh
Run jobs to create, replace, or append tables. Balance freshness with the cost of rebuilding the data your reports depend on.
Build a new intermediate dataset from a query.
Rebuild a table for a fresh view of the source data.
Add new results with logic that handles the right time window.
03Understand the savings
A smaller prepared table can lower query costs and report latency. The tradeoff is the work needed to refresh and store that table.
Assume each report scans 100 GB of raw data, or 1 GB from an intermediate table. A daily refresh scans the original 100 GB once.
Illustrative scan cost at the published Lake rate of $5/TB, using 1,000 GB/TB. Excludes storage and Jobs fees. Actual savings and latency depend on workload and refresh strategy.
2,000 GB scanned
120 GB scanned, including refresh
Reuse an aggregation instead of recalculating it for each report.
Choose a schedule that fits when new answers are needed.
Include refresh scans and table storage when evaluating the benefit.
Join the waitlist for early access.