Hi 5UDO,

How are you doing today?, As per my understanding, You're already doing a great job exploring advanced optimization techniques like partitioning, Z-Ordering, and even hashing, which shows you're thinking in the right direction. The fact that you’re close to your 200ms goal with around 270ms on a small warehouse is actually impressive given the size of your dataset. One suggestion would be to try using a larger warehouse temporarily to see how much that impacts latency, especially since Databricks SQL Warehouses are elastic and scale linearly for point lookups. Also, make sure that column stats are up to date by running ANALYZE TABLE COMPUTE STATISTICS FOR ALL COLUMNS—this helps with better query planning. Another angle is to materialize a lightweight index table with pre-hashed or encoded keys just for lookup purposes, or even cache hot subsets if the same queries run often. For more structured learning, Databricks’ official docs have a section on Delta Lake Performance Tuning, and the Databricks Academy also offers free and paid courses that cover best practices. Let me know if you want help designing a query strategy based on your access pattern!

Regards,

Brahma