R0 — Databricks Services & Terms: The Complete Glossary

Why this module exists

There are two moments this module is built for.

The night before an interview. You know the concepts. What you don't have is instant recall on which product is which, what got renamed, and which of the four things with "Lakeflow" in the name does what. That's a lookup problem, and lookups are what glossaries are for.

The meeting where someone says a product name you don't know. Somebody says "we'll put it behind a Genie Agent and expose it through Unity AI Gateway," and you have four seconds to decide whether to nod or ask. This module means you nod, correctly.

It's also designed to be listened to end to end — a Databricks A-to-Z for a commute. The six lessons are themed, not alphabetical, because a themed tour is something you can follow with your eyes on the road and an alphabetical list is not.

What makes this different from Databricks' own glossary

Three things.

Every entry says what people get wrong. A definition you can read off a marketing page is worth very little. "Auto Loader is not an alternative to Lakeflow pipelines, it runs inside one" is worth a lot.
Every entry carries release status where the docs state one — GA, Public Preview, Beta — and says "status not stated on the page" where they don't. A surprising number of things you'll hear pitched confidently in a meeting are Beta.
Old names are marked as old names. Eight renames landed between June 2025 and July 2026. This is the module that tells you Vector Search and AI Search are the same product.

How the entries are formatted

Every term uses the same four-line shape:

Term — one-sentence definition. · For: what you'd actually use it for. · Gotcha: the misconception, or the thing that bites. · Status: release status if the docs state one, plus the source and its last-updated date.

Read it left to right when you're studying. Skim only the bold terms when you're looking something up.

Lessons

#	Lesson	Terms	Read	Listen
1	Platform and compute	26	18 min	11 min
2	Storage and table formats	27	18 min	12 min
3	Ingestion and pipelines	27	19 min	13 min
4	Governance and sharing	31	19 min	13 min
5	AI, BI and apps	24	18 min	12 min
6	Dev tools, ops and 2026 arrivals	22	18 min	13 min

Then: Cheat sheet · Lab · Quiz

157 terms. The cheat sheet is all of them, one line each, on one page. It is the densest artifact in this course and it is the thing to print.

The four renames that matter most

If you retain nothing else from this module, retain these. Each one dates any material that uses the old name (full map in E0 lesson 5):

Say this	Not this	Changed
Lakeflow pipelines	Delta Live Tables, DLT	Jun 2025
OpenSharing	Delta Sharing	2026-06-10
AI Search	Vector Search	2026-06-01
Genie Agents · Genie Code · Genie One	Genie Spaces · Databricks Assistant · Databricks One	2026

Honest caveats

Three things I could not fully verify, flagged here rather than smoothed over:

Reyden — the engine name behind Lakehouse Real-Time appears in Databricks' newsroom and blog, not on the docs.databricks.com page for the product. The docs page describes Lakehouse//RT without naming the engine.
LTAP — announced 2026-06-16 in a press release. I could not find a docs.databricks.com page defining it. Treat it as a marketing-architecture term, not a product you provision.
"Budget policies" vs "serverless usage policies" — the AWS docs page has been retitled to serverless usage policies; the GCP page for the same feature still says serverless budget policies. The URL slug is still budget-policies on both. Live inconsistency, flagged in lesson 6.

Facts verified 2026-07-31 against the pages cited in each lesson.