Snowflake · SF-DS · Advanced

SnowPro Advanced: Data Scientist (DSA-C03) — Practice Questions and Mock Exam

Name: Pass-IT
Author: Pass-IT

Prepare for SF-DS with original practice questions and clear answer explanations. Ask Alex, your AI tutor, when you need more detail, use your results to identify topics to review, and practice your pacing with timed mock exams.

Start practicing SF-DS free Buy SF-DS access for $29.99

Buy SF-DS access for $29.99 Start practicing SF-DS free

65Mock exam questions

115minTime limit

750/ 1000Passing score

Checked against Snowflake · August 2026 · Current exam version

Overview

About the exam

The SnowPro Advanced: Data Scientist Certification (DSA-C03) validates the ability to apply data science principles within Snowflake, including defining data science concepts, performing data and feature engineering, building and deploying machine learning models, leveraging Snowpark for ML workflows, and presenting results through data visualization. It covers the complete model lifecycle from data preparation to production deployment.

This certification is for data scientists, ML engineers, and quantitative researchers with two or more years of practical data science experience with Snowflake in an enterprise environment. It demonstrates proficiency in building end-to-end ML solutions natively on Snowflake.

Try the practice preview

Try five SF-DS questions

Try five practice questions from the app’s current SnowPro Advanced: Data Scientist (DSA-C03) question bank, with answers and explanations.

Model Deployment1 / 5

When deploying a model via the Snowflake Model Registry, which method runs inference on a DataFrame?

AlexFull explanation from Alex

The ModelVersion.run() method is the correct API for running inference on a DataFrame through the Snowflake Model Registry. After retrieving a model version via Registry.get_model().version(), calling run(df) executes the model's prediction logic and returns results as a DataFrame. predict() resembles scikit-learn's API but is not a ModelVersion method. score() and execute() are also not valid ModelVersion methods. The run() method abstracts away model deserialization and provides a unified inference interface regardless of the underlying ML framework (scikit-learn, XGBoost, PyTorch, etc.).

Sourcedocs.snowflake.com

Model Development3 / 5

What is the primary purpose of the Snowflake Model Registry?

AlexFull explanation from Alex

The Snowflake Model Registry provides a centralized repository for logging, versioning, and managing the lifecycle of ML models as first-class schema-level objects. It supports model metadata, metrics tracking, version comparison, and deployment. The option “To store raw training data” is wrong: the registry stores models, not raw training data—use stages or tables for that. The option “To schedule data ingestion pipelines” is wrong: data pipelines are managed by Tasks and Streams, not the registry. The option “To monitor warehouse utilization” is wrong: warehouse monitoring uses ACCOUNT_USAGE schema and Resource Monitors. The registry integrates with scikit-learn, XGBoost, PyTorch, TensorFlow, and Hugging Face models. Any schema can serve as a registry without initialization.

Sourcedocs.snowflake.com

Data Science Concepts4 / 5

Which of the following best describes the F1 score?

AlexFull explanation from Alex

The F1 score is the harmonic mean of precision and recall: F1 = 2 * (P * R) / (P + R). The harmonic mean penalizes extreme imbalance—both precision and recall must be high for a high F1. The option “The arithmetic mean of precision and recall” (arithmetic mean) would be (P + R) / 2, which doesn't penalize imbalance as strongly and would overestimate performance. The option “The weighted average of all classification metrics” (weighted average of all classification metrics) is too broad and not the definition of F1. The option “The geometric mean of accuracy and recall” (geometric mean of accuracy and recall) uses the wrong metrics and wrong aggregation function. F1 is particularly useful for imbalanced classes where accuracy alone is misleading. Precision = TP/(TP+FP), Recall = TP/(TP+FN).

Sourcedocs.snowflake.com

Create a free account

Question pool

360 practice questions

The Pass-IT question pool gives you material to practice for SF-DS. A Pass-IT mock exam uses 65 questions and a 115-minute time limit; these are practice settings.

Pool details: SF-DS

Passing score750 / 1,000

as published by Snowflake

Objectives in the guide16 objectives listed in the official guide

across 4 domains in the official exam guide

Pool size360 questions

= The pool size is equivalent to 5 sets of 65 questions; this does not mean that each mock exam uses a separate set.

Blueprint domains4 domains in the exam blueprint

Data Science Concepts 61 · Data Preparation and Feature Engineering 99 · Model Development 106 · Model Deployment 94

Recorded as checked against sources360 of 360

questions recorded as having their answer, options, and explanation checked against official Snowflake documentation

How questions are made →

Exam domains

What's on the exam

Model Development carries the most weight at 31%, testing Snowpark ML connections, Cortex LLM usage, model training, validation, and interpretation. Data Preparation and Feature Engineering follows at 27%, covering cleaning, exploratory analysis, and feature transformations using Snowpark and SQL. Model Deployment sits at 25%, testing production moves, drift monitoring, and the Snowflake Model Registry.

Data Science Concepts is the lightest domain at 17%, covering ML fundamentals and statistics rather than Snowflake-specific tooling. Development, preparation, and deployment together make up 83% of the exam, so most of the value is in knowing how Snowflake-native tools handle each stage of the ML lifecycle rather than in general data science theory.

Exam blueprint: SF-DS

Data Science Concepts17%

Apply data science methodologies, statistics, and ML fundamentals within the Snowflake ecosystem.

≈ 14 h

Data Preparation and Feature Engineering27%

Prepare data, engineer features, and handle data quality using Snowflake SQL, Snowpark, and ML functions.

≈ 22 h

Model Development31%

Build, train, and evaluate ML models using Snowpark ML, Snowflake ML functions, and integrated frameworks.

≈ 25 h

Model Deployment25%

Deploy, monitor, and manage ML models in production using Snowflake model registry and serving infrastructure.

≈ 20 h

Format

Exam format and question types

The exam consists of 65 questions in 115 minutes, drawn from multiple-choice, multiple-select, and interactive formats. Most items present a data science problem (a dataset, a model choice, a deployment decision) and ask you to select the Snowflake-native method that fits it: Snowpark, Cortex, or the Model Registry. At roughly 1.8 minutes per question, the pace is workable if you're not deriving ML fundamentals from scratch.

Question types: SF-DS

Multiple Choice70%

Select the single answer that best meets the question’s requirements.

Multiple Response30%

Select multiple answers. Follow the question’s instructions on how many to choose.

See Snowflake for official question-format information. The shares shown describe the Pass-IT practice pool; they do not establish the proportions on the official exam.

Planning

Preparing for SF-DS

Delivered by online proctoring or at an onsite testing center, in English. The certification expires two years after your issue date; you recertify through the Snowflake Continuing Education program with an eligible instructor-led training course or an equivalent or higher-level SnowPro certification.

Preparation and logistics: SF-DS

Preparation

Illustrative study time50–120 h

illustrative planning range: 50 h with relevant experience to 120 h when starting out; your needs may fall outside this range

LevelAdvanced

Recommended backgroundSnowPro Core Certified. 2 or more years of hands-on Snowflake data science experience in production.

Taking and maintaining the certification

DeliveryOnline proctored or onsite testing centers.

Retake policyLimit of 4 attempts in a 12-month period. After three attempts Snowflake recommends attending an onsite Snowflake training course. Each registration requires the full registration fee.

Certification validity2 years

Snowflake certifications expire two years after the certification issue date. Recertify through the Snowflake Continuing Education (CE) program: complete an eligible Snowflake Instructor-Led (ILT) training course, or earn an equivalent or higher-level SnowPro certification. A valid certification is required to take part in the CE program.

Official guideView on provider site

Study notes

Common pitfalls

Topics to review: SF-DS

01Snowpark ML vs External
Not knowing when to use Snowpark ML (native, runs in Snowflake) versus external ML frameworks with Snowflake as a data source leads to wrong architecture answers.
02Feature Store
Confusing Snowflake feature store capabilities with external feature stores and not understanding how to manage feature pipelines natively leads to engineering question errors.
03Model Registry
Not understanding how Snowflake Model Registry works for versioning, deploying, and managing ML models leads to incorrect lifecycle management answers.
04UDFs for Inference
Misunderstanding how to use Python UDFs and vectorized UDFs for model inference at scale within Snowflake causes deployment pattern mistakes.
05Data Leakage
Not recognizing common data leakage patterns in feature engineering — such as using future data or target encoding without proper splits — leads to wrong methodology answers.

Practice with questions and explanations in Pass-IT →

Questions

Frequently asked questions

How long is the SnowPro Advanced: Data Scientist (DSA-C03) exam?

The SnowPro Advanced: Data Scientist (DSA-C03) exam has 65 questions and a 115-minute time limit.

What is the passing score for SnowPro Advanced: Data Scientist (DSA-C03)?

The passing score for the SnowPro Advanced: Data Scientist (DSA-C03) exam is 750 / 1000.

Which pitfalls should I review when preparing for SnowPro Advanced: Data Scientist (DSA-C03)?

Topics to review include Snowpark ML vs External, Feature Store, Model Registry, UDFs for Inference, Data Leakage. Work through examples to check that you understand the distinctions and can explain your answer.

How is the Advanced Data Scientist exam weighted?

Model development is the largest section at 31%, with data preparation and feature engineering at 27% and model deployment at 25%. Data science concepts take the remaining 17%. Deployment carrying a quarter is the part people underestimate, because it is Snowflake-specific rather than general practice.

What do you need before the Advanced Data Scientist exam?

SnowPro Core certification, plus two or more years of hands-on Snowflake data science work in production according to Snowflake's own guidance. The catalog budget is around 80 hours. The Core requirement is enforced, so it has to be passed first regardless of how much modelling experience you bring.

Advanced Data Scientist or Advanced Data Engineer?

The data scientist exam puts 31% on model development and 25% on deployment, so it is about building and shipping models inside Snowflake. The data engineer exam puts 28% on data movement and 25% on transformation, so it is about the pipelines that feed them. Both require SnowPro Core and neither ranks above the other.

How long does the Advanced Data Scientist certification last?

Two years from the issue date. You recertify through the continuing-education programme, using an eligible instructor-led course or an equivalent or higher SnowPro certification rather than a resit.

One certification, 12 months

Practice for SF-DS

Focus your practice on one certification, or choose Pro to practice across all certifications.

Buy SF-DS access for $29.99

Start a free practice sessionTry the first 20 questions without a card to see whether the practice suits you.

For eligible purchases: money-back guarantee if you fail your exam.

View guarantee terms →