GES-C01
Free SnowPro Specialty: Gen AI Practice Test
12 exam-style questions with full explanations — no sign-up. Score yourself, then close your gaps with the full course.
Exam fee ~$2255 exam domainsLevel Intermediate to Advanced2 timed practice tests in the course
✅ Free practice — no sign-up📝 Real exam-style questions💡 Detailed explanations💸 30-day money-back via Udemy
A support operations team wants to route thousands of incoming tickets per day into a fixed set of categories (Billing, Technical, Shipping, Other) without training a custom model or writing category-specific prompts for each ticket. Which Snowflake Cortex approach BEST meets this requirement?
CLASSIFY_TEXT is the purpose-built Cortex function for zero-shot text classification: you supply the text and a list of candidate labels and it returns the best-matching category, with no prompt engineering or training required.
A retail company wants business analysts who don't know SQL to ask questions like 'What were total sales by region last quarter?' directly against a Snowflake table containing verified sales data, and get accurate, governed SQL-backed answers. Which Cortex AI capability BEST meets this requirement?
Cortex Analyst is purpose-built for natural-language-to-SQL over structured data. It uses a semantic model (mapping business terms, synonyms, dimensions, and measures to real tables/columns) to reliably generate governed, accurate SQL from plain-English questions.
You are ranking retrieved passages for a RAG pipeline using two VECTOR(FLOAT, 768) columns. Which Snowflake function should you use so that a HIGHER returned value always means the two vectors are MORE similar?
VECTOR_COSINE_SIMILARITY returns a value between -1 and 1, where a higher value indicates greater similarity, making it directly usable for ORDER BY ... DESC ranking in retrieval.
An organization must fine-tune an open-source 7B-parameter LLM on proprietary support-ticket text while keeping all data and compute inside Snowflake's governance boundary, and the workload requires dedicated GPU acceleration for training. Which approach BEST meets these requirements?
Snowpark Container Services supports GPU-enabled compute pools (e.g. GPU_NV_S/M/L instance families) that let you run arbitrary custom containers, including training/fine-tuning jobs on open-source checkpoints, entirely within Snowflake's account boundary so data never leaves governance controls.
A Snowflake account administrator wants to grant a subset of roles the ability to call SNOWFLAKE.CORTEX.COMPLETE and other Cortex LLM functions, while preventing all other roles from using them. Which approach correctly implements this access control?
Access to Cortex LLM functions is controlled by granting the built-in SNOWFLAKE.CORTEX_USER database role to the roles that should be permitted to call functions like COMPLETE, SUMMARIZE, and TRANSLATE.
A developer calls SNOWFLAKE.CORTEX.COMPLETE with a smaller-context-window model and passes in a very long legal document plus a question about it. The call fails because the combined prompt and expected response exceed the model's context window. What is the BEST fix?
Context window limits are a hard token-count constraint that vary by model (e.g., some Cortex-hosted models support far larger windows than others). The correct fix is either choosing a model with a larger window or reducing the effective input size via chunking, summarization, or retrieval (e.g., Cortex Search) before generation.
An organization has 50,000 unstructured PDF legal contracts stored in Snowflake stages. They want to build a chatbot that retrieves relevant contract passages and lets an LLM answer employee questions using retrieval-augmented generation (RAG). Which Cortex AI service should be used to build the retrieval layer?
Cortex Search is Snowflake's managed hybrid search service purpose-built for RAG — it indexes unstructured text/document content and returns the most relevant passages to feed into an LLM prompt for grounded answers.
Before generating embeddings for a knowledge base of long support articles, a developer needs to split each article into smaller overlapping passages of a configurable size. Which Cortex function should be used?
SPLIT_TEXT_RECURSIVE_CHARACTER is the Cortex function designed to chunk text into smaller passages with configurable chunk size and overlap, which is exactly what's needed before embedding for retrieval.
A team wants to deploy an interactive chatbot UI backed by Cortex Search and Cortex LLM functions. The UI must run entirely inside Snowflake's security perimeter, be shareable through Snowsight via role-based access, and require no separate web server, hosting account, or ingress configuration to manage. Which deployment option BEST fits?
Streamlit in Snowflake apps run natively on a Snowflake warehouse, require no external hosting or ingress setup, and are shared to other users purely through Snowflake role grants and Snowsight — exactly matching the stated requirements.
A Snowflake admin needs to identify which roles and warehouses are consuming the most credits from Cortex LLM function calls (such as COMPLETE and SUMMARIZE) over the past 30 days. Which SNOWFLAKE.ACCOUNT_USAGE view should they query?
CORTEX_FUNCTIONS_USAGE_HISTORY records credit and token consumption for Cortex LLM function calls, broken down by function name, model, and consumer, making it the correct view for this monitoring task.
You need to generate 1024-dimensional vector embeddings for English product descriptions so you can build a custom similarity-search table using VECTOR data types and cosine distance. Which Cortex function should you call?
EMBED_TEXT_1024 generates 1024-dimensional embedding vectors using a supported embedding model, matching the dimensionality needed for the VECTOR column and distance comparison described.
A support team wants to automatically flag customer support tickets with a numeric score indicating how positive or negative the customer's tone is, without building a custom model. Which Cortex LLM function should they call?
SENTIMENT is a built-in Cortex LLM function that returns a numeric polarity score (negative to positive) for a piece of text, exactly matching the requirement with no custom model needed.
Ready for the real thing?
The full course has two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed answer explanations.
Start my full course on Udemy →GES-C01 exam — quick answers
How much does the GES-C01 exam cost?
The exam fee is approximately $225 and varies by region — confirm current pricing with the certification vendor before you book.
What topics are on the exam?
It covers 5 domains: Snowflake Gen AI and LLM Concepts (Objective group (confirm official Snowflake exam-guide weights)), Snowflake Cortex AI Features (Objective group (confirm official Snowflake exam-guide weights)), Working with Data for Gen AI (Objective group (confirm official Snowflake exam-guide weights)), Building and Deploying Gen AI Applications (Objective group (confirm official Snowflake exam-guide weights)), Governance, Security, and Cost (Objective group (confirm official Snowflake exam-guide weights)). The full course has a dedicated chapter, lab and practice-test coverage for each.
Is this practice test really free?
Yes — all questions on this page are free with explanations and no sign-up. The paid Udemy course adds two full-length timed exams, video lessons and hands-on labs.
Will this prepare me for the real exam?
The questions mirror the real exam's style and are mapped to the official domains. This is exam-focused preparation — combine the free test with the full course's timed simulations to gauge your readiness.