| Exam Code | GES-C02 |
| Exam Name | SnowPro Specialty Gen AI |
| Vendor | Snowflake |
| Total Questions | 52 |
| Last Updated | 30-Oct, 2026 |
- Latest & Verified Questions
- Detailed Explanations
- Regular Updates
- Full-Length Practice Tests
- Boost Your Exam Confidence
PDF Study Materials
Downloadable guides and resources
Practice Tests
Full-length tests with real exam experience
Progress Tracking
Monitor your performance and improvement
Study Anywhere
Access on desktop, tablet, and mobile
Free GES-C02 Practice Questions
Explore 5 free practice questions for the Snowflake GES-C02 exam and review the answers and explanations to support your preparation.
-
Question 1
Single answerA table claims.raw_extract stores the output of AI_EXTRACT, called without scores, in a VARIANT column ex. One field in the responseFormat was named policy_number. An analyst is writing a view that needs that value as a string column.
Which expression returns it?
-
Option A:Correct
ex:policy_number::STRING
-
Option B:Correct
ex:scoring:scores:policy_number::STRING
-
Option C:Correct
ex:response[0]:policy_number::STRING
-
Option D:Correct
ex:response:policy_number::STRING
Show Answer & Explanation Hide Answer & Explanation
Correct Answer: D
-
-
Question 2
Single answerA hotel group exposes its booking data to Cortex Analyst through a semantic view. The net_room_revenue metric is defined and calculates correctly, but revenue managers ask about "rooms income" and "lodging takings", and Analyst often fails to connect those phrases to the metric.
Which change to the semantic view most directly addresses this?
-
Option A:Correct
Add the managers' phrases as synonyms on that metric
-
Option B:Correct
Redefine net_room_revenue as a fact so that it is evaluated at row level instead of aggregated
-
Option C:Correct
Add a filter to the bookings logical table that keeps only completed and paid stays
-
Option D:Correct
Add the managers' phrases as sample_values on the booking date time dimension
Show Answer & Explanation Hide Answer & Explanation
Correct Answer: A
-
-
Question 3
Single answerA consumer electronics retailer must pick one of two Cortex-hosted models for drafting replies to product questions within the week. The reviewers are merchandising analysts who do not write code. They want to see both models answer the same sample questions side by side, try different temperature settings with Cortex Guard switched on, and then hand engineers the chosen configuration as SQL. No labeled set of approved replies exists yet.
What should the team use?
-
Option A:Correct
Set up AI Observability evaluation runs for both models and score their correctness against ground-truth replies
-
Option B:Correct
Compare both models side by side in Cortex Playground, then export the SQL
-
Option C:Correct
Have analysts edit a shared Snowflake Notebook that loops both models over the questions with AI_COMPLETE
-
Option D:Correct
Ask engineers to build a Streamlit in Snowflake app showing both replies
Show Answer & Explanation Hide Answer & Explanation
Correct Answer: B
-
-
Question 4
Single answerA rail operator's track inspectors complete paper forms that combine printed labels, tick boxes and handwritten remarks, and the forms are scanned to PDF. The operations lead assumes staff must first retype each form into a digital template before any Snowflake document function can use it.
-
Option A:Correct
Tick boxes are read from scans, but handwriting is not
-
Option B:Correct
Scanned files are accepted once the stage uses server-side encryption, which enables OCR on them
-
Option C:Correct
AI_EXTRACT reads checkboxes and handwriting directly from scanned files
-
Option D:Correct
Handwritten remarks are read only after AI_TRANSLATE has converted them into typed text first
Show Answer & Explanation Hide Answer & Explanation
Correct Answer: C
-
-
Question 5
Multiple answersA freight forwarder has an existing Cortex Agent, SHIPMENT_HELPER, owned by the AI_ENG role. Its tools are a Cortex Search service over customs notes and a semantic view over shipment tables. Dispatchers, whose default role is DISPATCH, must be able to ask the agent questions.
DISPATCH already holds the SNOWFLAKE.CORTEX_AGENT_USER database role, USAGE on the dispatchers' default warehouse and on the relevant databases and schemas, and everything the semantic view tool requires.
Which grants does DISPATCH still need?
(Choose two.)
-
Option A:Correct
USAGE on the SHIPMENT_HELPER agent itself
-
Option B:Correct
MODIFY on the SHIPMENT_HELPER agent
-
Option C:Correct
USAGE on the warehouse the AI_ENG team uses for its own work
-
Option D:Correct
USAGE on the agent's Cortex Search service
Show Answer & Explanation Hide Answer & Explanation
Correct Answer: A, D
-