Compile & DAG view: compare model SQL vs compiled SQL in-UI

Описание: Snowflake just made modeling a lot simpler. In this talk/demo, we walk through dbt Projects on Snowflake—a new capability that lets you run open-source dbt Core inside Snowflake, right next to your governed data. That means one UI, one security model, Snowflake warehouses for execution, built-in scheduling with Snowflake Tasks, first-class Git integration, and end-to-end observability.

If you already keep your data in Snowflake and need the T in ELT, this is a clean way to onboard teams to dbt without managing extra infra (no separate runners, no airflow required).

⏱️ Chapters

00:00 Screen share & setup hiccup (keepin’ it real)
00:25 What is “dbt Projects on Snowflake”? (dbt Core, inside Snowflake)
01:12 Why run dbt in Snowflake: simpler onboarding, governance, debugging
02:05 Self-service projects: create new or connect an existing Git repo
02:42 How it’s modeled: a new schema-level object for dbt projects
03:18 Workspaces IDE tour: models, tests, packages, profiles
04:10 Environments & profiles (dev/test/prod) without storing user creds
04:42 Compile & DAG view: compare model SQL vs compiled SQL in-UI
05:28 Managing dependencies (dbt deps) and dbt packages in a secure env
06:10 Scheduling with Snowflake Tasks (cron or UI)—no extra runner
07:00 Monitoring runs: logs, query IDs, memory usage, timings, lineage
07:42 Git workflows in the browser (branches, commits, push from UI)
08:25 Demo: run on dev warehouse, observe logs & Account Usage data
09:10 Use cases & scope: best for ELT when data already lives in Snowflake
09:48 Positioning vs dbt Cloud / external orchestration
10:20 Bonus: account-level observability & cost transparency
10:52 Links, sample repo (Tasty Bites), next steps & user group info

What you’ll learn
How dbt Core runs natively in Snowflake (governed, secure, observable)
Creating a project from scratch or connecting your GitHub repo
Using the Workspaces IDE to edit YAML/SQL/Python and view the DAG
Running compile, deps, test, and run directly in the UI
Scheduling jobs via Snowflake Tasks (UI or cron) with no extra infra
Monitoring: dbt logs, query IDs, timings, and Account Usage insights
Best practices for dev/test/prod profiles and warehouse selection
When to pick dbt Projects on Snowflake vs dbt Cloud / external runners

Who is this for?
Analytics engineers, data engineers, and BI teams who already load data into Snowflake and want a secure, low-ops path to production dbt with first-party scheduling and monitoring.

👉 Tell me in the comments: what’s your biggest blocker to shipping dbt models today—env management, scheduling, or debugging? I’ll cover the top issues in a follow-up.