The Work Gym

Courses / Data and analytics tools

Snowflake SQL

Snowflake SQL in a Snowsight-style worksheet: roles and warehouses, semi-structured data, time travel and more.

Start lesson 1 Practice interview Take the assessmentPractice the first unit of every course. No account needed.

63lessons

548checked exercises

12projects, cases and exams

Prepares you for SnowPro Core

Projects, case studies and mock exams

Project · IntermediateThe June web funnel

Turn a month of raw clickstream JSON into a conversion funnel by device, step by step.

6 steps · 35 min Project · IntermediateSupplier coverage audit

Build and publish a supplier coverage report without dropping missing supplier matches.

6 steps · 35 min Project · IntermediateA checked monthly close

Build a six-month report with explicit status rules and publish a secure view.

6 steps · 35 min Project · AdvancedCustomer profile release

Stage profile updates on a clone, audit the result and publish a secure summary view.

6 steps · 40 min Case study · AdvancedA store returns review

Identify the store to review using a defined order denominator and an observed return rate.

6 steps · 35 min Project · ExpertFrom raw JSON to a clean table

Turn landed JSON orders into a typed, one-row-per-line table, then keep it current with MERGE as new files arrive.

6 steps · 40 min Case study · ExpertThe returns question

Work out where returns come from, by month, channel, category and discount, and say what the numbers do and don't show.

6 steps · 30 min Mock exam · ExpertSnowPro Core COF-C02 format

One hundred original single-answer and multiple-response questions on the historical COF-C02 blueprint.

100 questions · 115 min Mock exam · ExpertSnowPro Core COF-C02 format

A second original 100-question exam following the historical COF-C02 domain weights and 115-minute format.

100 questions · 115 min

Practice Snowflake the way you'd use it at work: in a worksheet with a role, a warehouse and a database to run in. You start with the context and names that trip up newcomers, then Snowflake's own SQL: IFF, QUALIFY, GROUP BY ALL, LISTAGG and its date functions. From there you query JSON in VARIANT columns with paths and FLATTEN, build and change tables with MERGE, undo mistakes with time travel and UNDROP, and manage roles, grants and warehouses. Projects and interview sets build toward two timed exams using the historical COF-C02 blueprint. Every query is checked against a second copy of the data, so the rule you write has to be the right one.

Beginner

Getting around Snowflake

Check worksheet context, object names, roles and compute before running SQL.

5 lessons

1Worksheets and contextRun queries in a Snowsight worksheet, read the context it runs in, and name tables in full.5 exercises · 10 minFree

2Names, quotes and capital lettersHow Snowflake stores names in upper case, when double quotes make a name exact, and why data stays case-sensitive.5 exercises · 10 minFree

3Roles and warehousesSwitch roles to read what your current role can't, and pick the warehouse that runs your queries.5 exercises · 10 minFree

4Worksheet context checksCheck the active database, schema and role before you run a shared worksheet.4 exercises · 12 minFree

Practice setWorksheet setupResolve context, quoted names and permission problems in eight new worksheet tasks.Free

Query basics, the Snowflake way

Filter, group, join and work with dates using Snowflake expressions.

6 lessons

6Filter and sortKeep the rows you need with WHERE, IN, BETWEEN and ILIKE, then sort them, with NULLs where you want them.6 exercises · 12 minWith a plan

7IFF, NVL, DECODE and DIV0Snowflake's shorthand for if-then logic, missing values and safe division.5 exercises · 12 minWith a plan

8Totals by groupCount and total by group, with GROUP BY ALL, COUNT_IF, HAVING and LISTAGG.5 exercises · 12 minWith a plan

9Joins across tables and schemasJoin orders to customers, stores and products, find what has no match, and join across schemas.5 exercises · 14 minWith a plan

10Dates: DATEADD, DATEDIFF and DATE_TRUNCAdd to dates, measure the gap between them, round them to months, and format them, with Snowflake's date functions.5 exercises · 12 minWith a plan

Practice setSnowflake basicsTen mixed tasks on context, names, filters, shorthand functions, grouping, joins and dates.With a plan

Reliable relational queries

Handle missing values, clean text and preserve report grain.

4 lessons

12NULLs and report denominatorsDistinguish row counts, populated values and distinct values before calculating a rate.4 exercises · 13 minWith a plan

13Text keys and safe conversionsNormalize text keys, split email domains and keep missing data visible.4 exercises · 13 minWith a plan

14Join grain and unmatched rowsChoose the output grain before joining detail tables and preserve unmatched records deliberately.4 exercises · 15 minWith a plan

Practice setTrustworthy report queriesBuild eight reports on fresh filters while keeping NULLs, dates and detail grain correct.With a plan

Intermediate

Windows and report logic

Choose deterministic ranks, reuse CTEs and write checked report queries.

4 lessons

16QUALIFY and window functionsRank, number and total rows with window functions, and filter on them directly with QUALIFY.5 exercises · 14 minWith a plan

17Deterministic ranking with QUALIFYChoose between one row, tied ranks and distinct ranks, then filter the window result.4 exercises · 14 minWith a plan

18CTEs and viewsBreak a query into named steps with WITH, and save a query as a view for others to use.5 exercises · 12 minWith a plan

Practice setWindows and JSONEight mixed tasks on QUALIFY, window functions, VARIANT paths, FLATTEN and CTEs.With a plan

JSON and array grain

Parse documents, expand arrays and build a web-funnel report.

7 lessons

20JSON in VARIANT columnsRead nested JSON with paths, cast VARIANT values to real types, and handle keys whose type changes.6 exercises · 14 minWith a plan

21Parsing JSON text safelyTurn landed text into VARIANT, cast paths and isolate malformed records.4 exercises · 14 minWith a plan

22FLATTEN: one row per array elementTurn arrays and objects inside JSON into rows with LATERAL FLATTEN, keep empty arrays with OUTER, and flatten twice for nested data.5 exercises · 16 minWith a plan

23LATERAL arrays and document grainKeep document identifiers through FLATTEN and distinguish empty arrays from populated elements.4 exercises · 15 minWith a plan

24Build JSON: OBJECT_CONSTRUCT and ARRAY_AGGGo the other way: build objects and arrays from rows, parse JSON text, and test arrays with ARRAY_SIZE and ARRAY_CONTAINS.5 exercises · 12 minWith a plan

ProjectThe June web funnelTurn a month of raw clickstream JSON into a conversion funnel by device, step by step.With a plan

Practice setJSON report contractsCombine fresh JSON filters, deterministic windows and document-level checks.With a plan

Reporting projects

Audit supplier coverage and build a checked monthly close from detailed source rows.

4 lessons

ProjectSupplier coverage auditBuild and publish a supplier coverage report without dropping missing supplier matches.With a plan

ProjectA checked monthly closeBuild a six-month report with explicit status rules and publish a secure view.With a plan

Interview setEasy Snowflake questionsAnswer eight junior interview tasks with executable SQL and practical explanations.With a plan

Practice setReusable reporting queriesBuild eight fresh report requests with explicit dates, join grain and reusable definitions.With a plan

Advanced

Safe data changes

Validate change-feed grain and deliver an updated customer table.

5 lessons

31Create tablesCreate tables from a query or a column list, choose permanent, transient or temporary, and know which constraints Snowflake enforces.5 exercises · 14 minWith a plan

32INSERT, UPDATE, DELETE and MERGEChange rows safely, update one table from another, and upsert a change feed with MERGE.5 exercises · 16 minWith a plan

33MERGE source contractsAudit the source key, separate inserts from updates and apply a repeatable customer merge.4 exercises · 16 minWith a plan

ProjectCustomer profile releaseStage profile updates on a clone, audit the result and publish a secure summary view.With a plan

Practice setSafe data changesPractice independent table builds, merges, clones and repairs on new subsets.With a plan

Recovery and independent copies

Use historical reads, narrow repairs, UNDROP and CLONE to recover from mistakes.

4 lessons

36Time travel and UNDROPQuery a table as it was, find the statement that broke it, fix the data from history, and bring back a dropped table.6 exercises · 16 minWith a plan

37Zero-copy cloningCopy tables and whole schemas instantly with CLONE, including from a point in time, and know what a clone costs.5 exercises · 12 minWith a plan

38Recovery with a narrow repairInspect historical values, restore only damaged rows and retain a recovery clone.4 exercises · 15 minWith a plan

Practice setRecovery evidenceRead, compare and recover independently prepared incidents without replacing unaffected data.With a plan

Account controls and credits

Inspect metadata, apply least privilege and estimate compute consumption under stated assumptions.

7 lessons

40SHOW, DESCRIBE and INFORMATION_SCHEMAFind your way around an account with SHOW and DESCRIBE, query their output with RESULT_SCAN, and search INFORMATION_SCHEMA.5 exercises · 14 minWith a plan

41Roles, grants and the role hierarchyCreate a role, grant it what it needs to read a schema, fix a missing grant, and see how privileges flow up the hierarchy.6 exercises · 16 minWith a plan

42Warehouses and costCreate, resize and tune warehouses, and work out what queries cost in credits.6 exercises · 14 minWith a plan

43Permission checks with a working roleProve read access with the intended role and inspect the exact privileges it receives.4 exercises · 16 minWith a plan

44Warehouse credit estimatesEstimate compute credits from running time and distinguish estimates from account charges.4 exercises · 14 minWith a plan

Case studyA store returns reviewIdentify the store to review using a defined order denominator and an observed return rate.With a plan

Practice setAccount controlsCombine metadata, grants and warehouse settings in eight independent account tasks.With a plan

Expert

Files and incremental pipelines

Load staged fixtures, handle bad records and operate streams and explicit local tasks.

7 lessons

47Loading data: stages, COPY and SnowpipeHow files get into Snowflake: stages, file formats, COPY INTO and its options, and continuous loading with Snowpipe.5 exercises · 14 minWith a plan

48Stages and file loadingList staged files, load CSV and JSON, and inspect the rows that reached a table.4 exercises · 18 minWith a plan

49COPY errors and validationChoose row or file error handling, validate without loading, and read the load history.4 exercises · 18 minWith a plan

50Streams and consumptionRead change rows and consume a stream through a successful write.4 exercises · 19 minWith a plan

51Tasks and explicit runsCreate a suspended task, resume it, execute it and inspect its output.4 exercises · 18 minWith a plan

ProjectFrom raw JSON to a clean tableTurn landed JSON orders into a typed, one-row-per-line table, then keep it current with MERGE as new files arrive.With a plan

Practice setAccount operationsLoad new file selections, operate a task, consume a stream and test a governed report.With a plan

Governed reports and query profiles

Test policy branches, secure projections and conservative partition-pruning estimates.

5 lessons

54Performance: pruning, clustering and cachingHow Snowflake stores tables in micro-partitions, skips them with pruning, and answers repeat queries from its caches.6 exercises · 14 minWith a plan

55Query profiles and pruningCompare scanned partitions with total partitions and test selective filters.4 exercises · 18 minWith a plan

56Policies and secure viewsAttach masking and row access policies, then test the output as different roles.4 exercises · 22 minWith a plan

Case studyThe returns questionWork out where returns come from, by month, channel, category and discount, and say what the numbers do and don't show.With a plan

Practice setGoverned query resultsTest secure projections, policy branches and selective scans on new report requests.With a plan

Interviews and certification practice

Solve medium and hard interview tasks and two original timed COF-C02 practice exams.

5 lessons

Interview setMedium Snowflake questionsThe Snowflake questions data interviews ask: deduplication, JSON, time travel, table types, cloning and the architecture.With a plan

Interview setHard Snowflake questionsSolve eight senior tasks on deterministic updates, grain, JSON and account decisions.With a plan

Mock examSnowPro Core COF-C02 formatOne hundred original single-answer and multiple-response questions on the historical COF-C02 blueprint.With a plan

Mock examSnowPro Core COF-C02 formatA second original 100-question exam following the historical COF-C02 domain weights and 115-minute format.With a plan

Practice setFinal Snowflake reviewCombine six executable report and pipeline tasks with two operational decisions.With a plan