Snowflake SnowPro Specialty Gen AI GES-C01 Practice Test 52
Question 510
At a consumer travel platform, a Snowflake engineering team is undertaking a booking assistant redesign. A legal department wants temporary source PDFs removed after extracting permitted metadata. Budget and operational risk are both material; peak demand is highly seasonal; a bad response can trigger a financial adjustment. Which decision MOST appropriately balances the requested outcome with production constraints?
- Couple index membership and search filters to document ACLs.
- Define and enforce distinct retention for originals, derived text, embeddings, and backups.
- Use immutable release metadata with time-bounded logs.
- Separate informational output from executable action permissions.
- Track retention obligations across every derived artifact.
Explanation
Correct Answers
B:This implements the response to the stated failure: Deleting the visible PDF alone may leave derived sensitive text indexed. It must be validated using verify deletion propagation and lineage These author-intended interpretations require independent technical review before operational use.
E:This establishes the ongoing safety boundary after the immediate correction. It provides protection while the team checks that expired content is removed from all applicable stores under policy. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This guardrail primarily addresses private document indexing and is not the companion intervention the case most urgently requires. The stated issue is a legal department wants temporary source pdfs removed after extracting permitted metadata. These author-intended interpretations require independent technical review before operational use.
C:This guardrail primarily addresses audit trail and model changes and is not the companion intervention the case most urgently requires. The stated issue is a legal department wants temporary source pdfs removed after extracting permitted metadata. These author-intended interpretations require independent technical review before operational use.
D:This guardrail primarily addresses llm output safety and is not the companion intervention the case most urgently requires. The stated issue is a legal department wants temporary source pdfs removed after extracting permitted metadata. These author-intended interpretations require independent technical review before operational use.
Question 511
At a national furniture retailer, a Snowflake engineering team is undertaking an automated customer feedback review. AI application users need to parse documents in a stage but should not alter the stage contents. Before implementation, promotion teams use rapidly changing product catalogs; managers want segment-specific reporting. Which first architectural or operating step MOST directly addresses this case?
- Grant narrowly scoped read access rather than broad stage-management privileges.
- Restrict source material and retrieval results to permitted audiences before deployment.
- Verify factual assertions against actual retrieved records before distributing the report.
- Check the appropriate Snowflake Cortex database-role grants and assign only the function privileges it needs.
Explanation
Correct Answer
A:The decision addresses the central constraint: ai application users need to parse documents in a stage but should not alter the stage contents. The most direct next step is to grant narrowly scoped read access rather than broad stage-management privileges. Before changing production, review both: test stage read versus write operations under the app role and inspect grant inheritance.. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:This can be appropriate when a shared search service indexes hr contracts that contain protected employee details. It addresses private document indexing, whereas the present limitation is better explained by excessive stage privileges increase tampering and disclosure risk. The target outcome here is that the app can process approved documents but cannot replace source files. These author-intended interpretations require independent technical review before operational use.
C:This can be appropriate when a generated audit summary cites a warehouse table that was never queried. It addresses model output hallucination, whereas the present limitation is better explained by excessive stage privileges increase tampering and disclosure risk. The target outcome here is that the app can process approved documents but cannot replace source files. These author-intended interpretations require independent technical review before operational use.
D:This can be appropriate when an application role can query sales tables but receives a permission denial for an ai inference function. It addresses cortex role grant, whereas the present limitation is better explained by excessive stage privileges increase tampering and disclosure risk. The target outcome here is that the app can process approved documents but cannot replace source files. These author-intended interpretations require independent technical review before operational use.
Question 512
At a regional logistics provider, a Snowflake engineering team is undertaking a warehouse dispatch modernization. AI application users need to parse documents in a stage but should not alter the stage contents. The proposed fix is still hypothetical; legacy barcode scanners cannot be replaced this quarter, and customer-facing delays must remain measurable. Which TWO checks provide the most decisive evidence before approving a change?
- Test stage read versus write operations under the app role
- Inspect downstream workflow authority
- Inspect grant inheritance in the affected workflow
- Compare each citation and metric to executed query output
- Review the requested command versus granted authority
Explanation
Correct Answers
A:This observation directly tests the case mechanism: Excessive stage privileges increase tampering and disclosure risk. In the described setting, it helps determine whether grant narrowly scoped read access rather than broad stage-management privileges. is justified. These author-intended interpretations require independent technical review before operational use.
C:This complementary observation tests a different part of the same hypothesis and should be read together with the first check before authorizing change. Expected success would be: the app can process approved documents but cannot replace source files. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:Checking this might help diagnose human-in-loop decisions, where a generative model proposes denying insurance claims without individual review. It is not one of the two most direct checks for the present case because the leading mechanism is excessive stage privileges increase tampering and disclosure risk. These author-intended interpretations require independent technical review before operational use.
D:Checking this might help diagnose model output hallucination, where a generated audit summary cites a warehouse table that was never queried. It is not one of the two most direct checks for the present case because the leading mechanism is excessive stage privileges increase tampering and disclosure risk. These author-intended interpretations require independent technical review before operational use.
E:Checking this might help diagnose llm output safety, where a support bot composes instructions that could expose or damage production data when followed. It is not one of the two most direct checks for the present case because the leading mechanism is excessive stage privileges increase tampering and disclosure risk. These author-intended interpretations require independent technical review before operational use.
Question 513
At a municipal utilities contractor, a Snowflake engineering team is undertaking a vendor compliance portal. AI application users need to parse documents in a stage but should not alter the stage contents. Diagnosis is disputed because document authors sometimes embed arbitrary instructions in descriptions; audit requirements prohibit silent data leakage. Which underlying failure mechanism BEST explains the observed limitation?
- Lowest per-token price may increase downstream errors and remediation spend.
- Warehouse masking on one output path does not guarantee all prompt-building logic excludes raw values.
- Aggregate performance hides concentrated failures for underrepresented users.
- Excessive stage privileges increase tampering and disclosure risk.
Explanation
Correct Answer
D:This is the causal hypothesis consistent with the observed limitation: ai application users need to parse documents in a stage but should not alter the stage contents. Confirm it through both test stage read versus write operations under the app role and inspect grant inheritance. before changing the service. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:That mechanism could explain a different case about inference cost versus quality, but does not most directly account for this specific symptom: ai application users need to parse documents in a stage but should not alter the stage contents. Here, the evidence to collect is test stage read versus write operations under the app role These author-intended interpretations require independent technical review before operational use.
B:That mechanism could explain a different case about data masking before prompts, but does not most directly account for this specific symptom: ai application users need to parse documents in a stage but should not alter the stage contents. Here, the evidence to collect is test stage read versus write operations under the app role These author-intended interpretations require independent technical review before operational use.
C:That mechanism could explain a different case about ai evaluation bias, but does not most directly account for this specific symptom: ai application users need to parse documents in a stage but should not alter the stage contents. Here, the evidence to collect is test stage read versus write operations under the app role These author-intended interpretations require independent technical review before operational use.
Question 514
At a professional services consultancy, a Snowflake engineering team is undertaking a role-scoped analytics portal. AI application users need to parse documents in a stage but should not alter the stage contents. The change was trialed on a small dataset; project billing must be attributable, while consultants work on multiple confidential clients. Which acceptance condition BEST demonstrates that the intervention worked?
- Clients cannot retrieve shared service secrets from the application.
- The app can process approved documents but cannot replace source files.
- Published internal summaries contain only traceable factual claims.
- Unused services stop accumulating unnecessary charges after approved suspension.
Explanation
Correct Answer
B:This is the observable acceptance condition that would show the intended fix addressed excessive stage privileges increase tampering and disclosure risk. It is stronger than activity counts because it tests the user-facing effect. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This could demonstrate success for security credential handling, but it would not establish resolution of the current issue: ai application users need to parse documents in a stage but should not alter the stage contents. Validate the app can process approved documents but cannot replace source files. instead. These author-intended interpretations require independent technical review before operational use.
C:This could demonstrate success for model output hallucination, but it would not establish resolution of the current issue: ai application users need to parse documents in a stage but should not alter the stage contents. Validate the app can process approved documents but cannot replace source files. instead. These author-intended interpretations require independent technical review before operational use.
D:This could demonstrate success for cortex search operating costs, but it would not establish resolution of the current issue: ai application users need to parse documents in a stage but should not alter the stage contents. Validate the app can process approved documents but cannot replace source files. instead. These author-intended interpretations require independent technical review before operational use.
Question 515
At an industrial maintenance company, a Snowflake engineering team is undertaking an asset fault knowledge base. AI application users need to parse documents in a stage but should not alter the stage contents. An auditor will inspect the control design; repair crews require concise results, and field logs contain abbreviations and exact part numbers. Which ongoing control MOST directly reduces the original risk?
- Integrate model allowlist changes into change control.
- Add adversarial retrieval cases to release validation.
- Inventory active services with accountable owners.
- Separate document ingestion and inference roles.
Explanation
Correct Answer
D:This retains a specific safeguard against recurrence of excessive stage privileges increase tampering and disclosure risk. It complements the immediate intervention, grant narrowly scoped read access rather than broad stage-management privileges., and allows the team to verify the intended result. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This is a sensible safeguard for model allowlist policy, but it does not target the concrete exposure in this scenario. The current failure is excessive stage privileges increase tampering and disclosure risk. These author-intended interpretations require independent technical review before operational use.
B:This is a sensible safeguard for prompt injection threat modeling, but it does not target the concrete exposure in this scenario. The current failure is excessive stage privileges increase tampering and disclosure risk. These author-intended interpretations require independent technical review before operational use.
C:This is a sensible safeguard for cortex search operating costs, but it does not target the concrete exposure in this scenario. The current failure is excessive stage privileges increase tampering and disclosure risk. These author-intended interpretations require independent technical review before operational use.
Question 516
At a renewable-energy service provider, a Snowflake engineering team is undertaking an engineering knowledge assistant. AI application users need to parse documents in a stage but should not alter the stage contents. Budget and operational risk are both material; technicians work in intermittent connectivity conditions; the business needs verifiable source references. Which decision MOST appropriately balances the requested outcome with production constraints?
- Grant narrowly scoped read access rather than broad stage-management privileges. Then separate document ingestion and inference roles.
- Check the appropriate Snowflake Cortex database-role grants and assign only the function privileges it needs. Postpone the critical evidence checks until after a broad release.
- Verify factual assertions against actual retrieved records before distributing the report. Postpone the critical evidence checks until after a broad release.
- Restrict source material and retrieval results to permitted audiences before deployment. Postpone the critical evidence checks until after a broad release.
Explanation
Correct Answer
A:This combines a corrective action with a bounded production safeguard, addressing ai application users need to parse documents in a stage but should not alter the stage contents. Prove the result by checking that the app can process approved documents but cannot replace source files. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:This proposal begins with an intervention for cortex role grant and postpones evidence that is essential for the current case. Before rollout, the team should grant narrowly scoped read access rather than broad stage-management privileges. These author-intended interpretations require independent technical review before operational use.
C:This proposal begins with an intervention for model output hallucination and postpones evidence that is essential for the current case. Before rollout, the team should grant narrowly scoped read access rather than broad stage-management privileges. These author-intended interpretations require independent technical review before operational use.
D:This proposal begins with an intervention for private document indexing and postpones evidence that is essential for the current case. Before rollout, the team should grant narrowly scoped read access rather than broad stage-management privileges. These author-intended interpretations require independent technical review before operational use.
Question 517
At a renewable-energy service provider, a Snowflake engineering team is undertaking an engineering knowledge assistant. A reviewer cannot reproduce why a classified record changed its label after a model refresh. Before implementation, technicians work in intermittent connectivity conditions; the business needs verifiable source references. Which first architectural or operating step MOST directly addresses this case?
- Apply log minimization and retention policies while preserving operational diagnostics.
- Record source revision, model identifier, prompt version, and time of each decision.
- Minimize and mask identifying fields before any generative AI invocation.
- Check the appropriate Snowflake Cortex database-role grants and assign only the function privileges it needs.
Explanation
Correct Answer
B:The decision addresses the central constraint: a reviewer cannot reproduce why a classified record changed its label after a model refresh. The most direct next step is to record source revision, model identifier, prompt version, and time of each decision. Before changing production, review both: replay a non-sensitive golden record and compare logged environment and predicted label.. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:This can be appropriate when a service traces entire user prompts and model outputs, accidentally retaining personal records for years. It addresses sensitive output logging, whereas the present limitation is better explained by missing version lineage prevents reliable fault attribution. The target outcome here is that every material label has enough provenance for reproducible review. These author-intended interpretations require independent technical review before operational use.
C:This can be appropriate when an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. It addresses data masking before prompts, whereas the present limitation is better explained by missing version lineage prevents reliable fault attribution. The target outcome here is that every material label has enough provenance for reproducible review. These author-intended interpretations require independent technical review before operational use.
D:This can be appropriate when an application role can query sales tables but receives a permission denial for an ai inference function. It addresses cortex role grant, whereas the present limitation is better explained by missing version lineage prevents reliable fault attribution. The target outcome here is that every material label has enough provenance for reproducible review. These author-intended interpretations require independent technical review before operational use.
Question 518
At a non-profit social services network, a Snowflake engineering team is undertaking a service quality dashboard. A reviewer cannot reproduce why a classified record changed its label after a model refresh. The proposed fix is still hypothetical; some participant data is confidential, and the organization serves populations with different access constraints. Which TWO checks provide the most decisive evidence before approving a change?
- Inspect active role and grants
- Compare logged environment and predicted label.
- Inspect actual request payloads in a safe test
- Replay a non-sensitive golden record
- Compare each citation and metric to executed query output
Explanation
Correct Answers
B:This complementary observation tests a different part of the same hypothesis and should be read together with the first check before authorizing change. Expected success would be: every material label has enough provenance for reproducible review. These author-intended interpretations require independent technical review before operational use.
D:This observation directly tests the case mechanism: Missing version lineage prevents reliable fault attribution. In the described setting, it helps determine whether record source revision, model identifier, prompt version, and time of each decision. is justified. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:Checking this might help diagnose cortex role grant, where an application role can query sales tables but receives a permission denial for an ai inference function. It is not one of the two most direct checks for the present case because the leading mechanism is missing version lineage prevents reliable fault attribution. These author-intended interpretations require independent technical review before operational use.
C:Checking this might help diagnose data masking before prompts, where an analysis app includes unmasked customer identifiers in an llm prompt unnecessarily. It is not one of the two most direct checks for the present case because the leading mechanism is missing version lineage prevents reliable fault attribution. These author-intended interpretations require independent technical review before operational use.
E:Checking this might help diagnose model output hallucination, where a generated audit summary cites a warehouse table that was never queried. It is not one of the two most direct checks for the present case because the leading mechanism is missing version lineage prevents reliable fault attribution. These author-intended interpretations require independent technical review before operational use.
Question 519
At a packaging manufacturer, a Snowflake engineering team is undertaking a product support search interface. A reviewer cannot reproduce why a classified record changed its label after a model refresh. Diagnosis is disputed because knowledge articles update every workday; operators need reliable revision dates. Which underlying failure mechanism BEST explains the observed limitation?
- Ordinary SELECT access on user tables does not automatically grant every managed AI capability.
- Fluent responses may invent provenance when source data is absent.
- Application convention is not the same as enforced model governance.
- Missing version lineage prevents reliable fault attribution.
Explanation
Correct Answer
D:This is the causal hypothesis consistent with the observed limitation: a reviewer cannot reproduce why a classified record changed its label after a model refresh. Confirm it through both replay a non-sensitive golden record and compare logged environment and predicted label. before changing the service. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
A:That mechanism could explain a different case about cortex role grant, but does not most directly account for this specific symptom: a reviewer cannot reproduce why a classified record changed its label after a model refresh. Here, the evidence to collect is replay a non-sensitive golden record These author-intended interpretations require independent technical review before operational use.
B:That mechanism could explain a different case about model output hallucination, but does not most directly account for this specific symptom: a reviewer cannot reproduce why a classified record changed its label after a model refresh. Here, the evidence to collect is replay a non-sensitive golden record These author-intended interpretations require independent technical review before operational use.
C:That mechanism could explain a different case about model allowlist policy, but does not most directly account for this specific symptom: a reviewer cannot reproduce why a classified record changed its label after a model refresh. Here, the evidence to collect is replay a non-sensitive golden record These author-intended interpretations require independent technical review before operational use.
Question 520
At a semiconductor equipment maker, a Snowflake engineering team is undertaking a component troubleshooting assistant. A reviewer cannot reproduce why a classified record changed its label after a model refresh. The change was trialed on a small dataset; engineers need verified part-specific guidance, while document titles are duplicated across product revisions. Which acceptance condition BEST demonstrates that the intervention worked?
- Every material label has enough provenance for reproducible review.
- Selected configuration satisfies defined quality and budget thresholds.
- No claim is denied solely on unreviewed generated text.
- The supported replacement meets functional and privacy requirements.
Explanation
Correct Answer
A:This is the observable acceptance condition that would show the intended fix addressed missing version lineage prevents reliable fault attribution. It is stronger than activity counts because it tests the user-facing effect. These author-intended interpretations require independent technical review before operational use.
Incorrect Answers
B:This could demonstrate success for inference cost versus quality, but it would not establish resolution of the current issue: a reviewer cannot reproduce why a classified record changed its label after a model refresh. Validate every material label has enough provenance for reproducible review. instead. These author-intended interpretations require independent technical review before operational use.
C:This could demonstrate success for human-in-loop decisions, but it would not establish resolution of the current issue: a reviewer cannot reproduce why a classified record changed its label after a model refresh. Validate every material label has enough provenance for reproducible review. instead. These author-intended interpretations require independent technical review before operational use.
D:This could demonstrate success for vendor feature deprecation, but it would not establish resolution of the current issue: a reviewer cannot reproduce why a classified record changed its label after a model refresh. Validate every material label has enough provenance for reproducible review. instead. These author-intended interpretations require independent technical review before operational use.