# Gilbert Databricks User Group > A free Databricks user group in Gilbert, Arizona, organised by TechFabric, with every session written up in full. Canonical: https://www.techfabric.com/user-groups/gilbert --- Gilbert , United States ## Gilbert Databricks User Group For data leaders and teams in the Phoenix metro building on Databricks, or deciding whether to. We host it at our office in Gilbert, twenty minutes from Tempe and Chandler. Clients bring TechFabric in when their reports don't agree or their AI never got past a pilot, and we fix the data underneath and put the AI into production. Each session builds one piece of that live, permissions included, so you see what it takes in your own workspace. Where it meets TechFabric, EZ Spaces Coworking, 1530 E Williams Field Rd, Ste 201, Gilbert, AZ 85295 Who runs it Preetham Reddy, founder and CEO of TechFabric, lead organizer What a session is like One thing built in the room, live, from a script rather than from screenshots, then questions for as long as people have them. Free, food provided, and no pitch. What the sessions cover - Lakebase and Postgres - Unity Catalog - Databricks Apps - Genie - AI agents - Lakeflow ### What the first session measured On 18 September 2026 Preetham Reddy, our founder and the chapter's lead organizer, built an application on Lakebase in the room, from a script, and every number on the slides came from that run rather than from a vendor deck. Branching the whole database reached READY in 4.5 seconds. A cold query took 2,979 ms against 1,650 ms warm. A synced table came online 51 seconds after the call, and writes from the app reached Delta in 10 to 15 seconds through the change data feed. The parts that fought back were the useful ones. History retention was seven days, and once that's gone there's no moment left to branch from, so it's a number to settle before somebody needs it. REPLICA IDENTITY FULL has to be set before the change data feed is turned on, because a table without it is skipped and flipping it afterwards isn't reliably picked up. And one row-level security policy on demo_app.orders gave three callers three different answers, down to a 403 with 42501 when a caller tried to write a row the policy would have hidden. The full write-up and all the slides are on the session's page, and the scripts are public at github.com/Fabric-Pro/fabric-databricks-usergroup , so you can run the same thing in your own workspace. ### Coming to a session - 01 RSVP where it's listed Sessions are posted on the chapter's page at usergroups.databricks.com , and sometimes cross-listed on our Phoenix Data & AI Meetup group . An RSVP on either one counts. - 02 Find Suite 201 We meet at our office inside EZ Spaces Coworking, 1530 E Williams Field Rd, Suite 201, Gilbert, about twenty minutes from Tempe and Chandler. The first session ran from 6 to 8pm Arizona time, which doesn't change with daylight saving. - 03 Bring a laptop if you want to follow along The build runs from a script, so you can run it beside us. At the first session a good number of people had laptops open and followed it as it ran. - 04 Stay for the questions They ran about half an hour last time, and they're where people bring problems from their own estates. There's no pitch at the end. ### Two chapters, one building The Phoenix Databricks User Group is a separate chapter with its own organisers, and it also meets at EZ Spaces on Williams Field Road. If a listing gives that address, check which chapter's page it's on before you assume it's this one. Sessions on this page, and on the chapter's own page at usergroups.databricks.com, are the Gilbert chapter's. #### If you'd like to give a talk here - Something you built, on Databricks, that is running somewhere - A script that builds it live, rather than screenshots of it working - At least one number you measured yourself - The part that fought back, and what you did about it - Which pieces are still in preview, said plainly ### Sessions so far Friday, September 18 #### Building applications on Lakebase: Postgres, Databricks Apps and AI in one place Preetham Reddy built an app live on Lakebase, with an AI agent that sees only what its user may see. It's how we'd give the business an app on its own data, with Postgres beside the analytics and no copy job between them. Preetham Reddy · 21 slides · 11 photos ### Give a talk here The sessions that go best are someone showing a thing they built and the parts that fought back. If you have one, tell us what you built and we will find you an evening.