kumyu.Learn
← Kumyu home日本語 ↗

DATA + AI / LEARNING PATHS

Databricks

Build a mental model of the platform, then choose what to study for your work or certification.

CORE LEARNING

Understand the platform, then its decisions.

Start with the platform map, then follow data from ingestion to governed use. AI, Lakebase, and Lakewatch are explained with their confirmed scope and availability.

System mapSetting → effectDesign decisionExercise & limits

Foundations course

Read in this order. Each article links to the previous and next chapter.

Published Updated
Intermediate

01. Databricks as a system: data, AI, Lakebase and Lakewatch

Follow a support request across governed analytical data, agent tools, transactional state and security telemetry. Learn which boundary each service owns and which settings change the design.

18 min↗
Published Updated
Foundation → design practice

02. Auto Loader: from arriving files to trustworthy BI

Connect file discovery, checkpoint state, schema decisions, Delta layers and BI grain; reason about settings, recovery, permissions and cost.

32 min↗
Published Updated
Foundation → design practice

03. Lakebase: separate serving data from transactional truth

Design a returns service that reads analytical features and writes operational decisions. Trace synced tables, Postgres transactions, change feeds and LTAP Direct Writes, then reason about keys, scaling, availability and cost.

27 min↗
Published Updated
Intermediate

04. Data-connected AI with Databricks: Agent Bricks and Unity Gateway in practice

Design authorization, meaning, audit, and cost before retrieval quality when connecting AI to enterprise data.

8 min↗
Published Updated
Intermediate → agent system design

05. Databricks AI roles: Genie, governed agents, memory and evaluation

Assign business questions, agent workflows and coding assistance to the right roles. Follow tool identities, state isolation, budget limits and failure evidence in a support case.

25 min↗
Published Updated
Foundation → security design

06. Lakewatch: follow security evidence from collection to response

Map security telemetry, normalization, governed evidence, detections and agent-assisted investigation. Use a synthetic AWS-oriented case to distinguish an alert, an incident decision and an executed response.

24 min↗

EXAM PREPARATION

Choose a certification to practise.

8 credentials · 391 original Japanese questions. Review each answer with explanations and official references. Check the current exam guide before taking an exam.

Independent learning material based on public sources. It is not an official question bank. Exercise code has not been run in Databricks; difficulty and passing results have not been calibrated.

Renewal practice

For existing credential holders. This 25-question set reuses Associate questions and is included in the total above.

RENEWAL

Data Engineer Associate Renewal

25 original Japanese questions · Answers, explanations, and official references

Official certification overview ↗

Articles cite official documentation, blogs, and videos when used. Community observations are labelled separately. Check each article’s review date, cloud scope, preview status, and exercise status before applying it.