O R A C L E C E R T I F I C A T I O N
1Z0-1110 Oracle Cloud Infrastructure Data Science 2022 Professional Practice Exam
Exam Number: 4824 | Last updated April 19, 2026 | 700+ questions across 4 vendor-aligned objectives
The 1Z0-1110 Oracle Cloud Infrastructure Data Science 2022 Professional exam is aimed at data scientists and machine-learning engineers who build, deploy, and operate models on OCI Data Science. Candidates validate command of notebook sessions, jobs, model deployment, the Accelerated Data Science library, and the integration points with Autonomous Database, Object Storage, and Data Catalog.
The heaviest content is Model Development and the ADS Library (roughly 35%), covering notebook sessions, conda environments, feature engineering with Accelerated Data Science, automated machine learning, and model registry publication. Jobs and Pipelines contributes another 20% with repeatable training runs and multi-step ML workflows.
Model Deployment and MLOps sits near 25% and drills into model deployment, logging, metrics, and traffic-splitting for safe rollout. Data Access, Governance, and Integration rounds out the remaining weight with Object Storage, Autonomous Database, Data Catalog, and the resource-principal model that authorizes service-to-service calls.
Every answer links to the source. Each explanation below includes a hyperlink to the exact Oracle documentation page the question was derived from. PowerKram is the only practice platform with source-verified explanations. Learn about our methodology →
608
practice exam users
97.0%
satisfied users
95.7%
passed the exam
4.2/5
quality rating
Test your 1Z0 1110 OCI Data Science Pro knowledge
10 of 700+ questions
Question #1 - Model Development and ADS
A data scientist at a credit-risk team is starting a new project on OCI Data Science. She needs a Jupyter environment with GPU access, a specific Python 3.9 stack with TensorFlow and pandas, and quick access to the team’s S3-compatible bucket.
Which OCI Data Science construct delivers this environment?
A) A notebook session with a GPU shape and a custom conda environment containing the required packages.
B) Oracle Functions with a Python runtime.
C) A manual Jupyter install on the scientist’s laptop.
D) A plain compute instance with no data science configuration.
Show solution
Correct answers: A – Explanation:
Notebook sessions in OCI Data Science provide GPU shapes and custom conda environments packaged with the needed libraries — the designed environment. Option D lacks the data science tooling. Option B is serverless, not interactive. Option C is not managed by Oracle. Source: Check Source
Question #2 - Model Development and ADS
A senior ML engineer wants to load a 4 GB CSV of transaction data into a notebook, profile it, and detect features that are highly correlated with the fraud label — without writing boilerplate pandas code.
Which OCI Data Science library accelerates this exploratory step?
A) Pure base pandas with no helper library.
B) Accelerated Data Science (ADS) library helpers for data loading, profiling, and feature correlation.
C) Excel opened on a laptop.
D) A Java program in Tomcat.
Show solution
Correct answers: B – Explanation:
ADS provides helpers for data ingestion, profiling (feature types, summary), and correlation analysis that abstract away boilerplate — designed for exactly this step. Option A works but is slower to iterate with. Option D is the wrong stack. Option C does not scale. Source: Check Source
Question #3 - Jobs and Pipelines
A team wants a reproducible training run: same script, same data, same environment, scheduled nightly. Outputs (model artifacts, metrics) must be captured for audit.
Which OCI Data Science feature fits a reproducible, scheduled training run with artifact logging?
A) Relying on the team to remember to run it.
B) Running the script manually in a notebook each night.
C) Data Science Jobs with a containerized or conda-based execution environment and artifact logging.
D) A shell script on the data scientist’s laptop.
Show solution
Correct answers: C – Explanation:
OCI Data Science Jobs are designed for repeatable, scheduled execution with environment control and artifact logging. Option B is manual. Option A is unreliable. Option D is off-platform. Source: Check Source
Question #4 - Model Deployment and MLOps
A product manager wants to release a new version of a fraud-detection model, routing 10% of predictions to the new version while keeping the other 90% on the current version to validate safely.
Which OCI Data Science model deployment feature supports this canary pattern?
A) Deploying the new version with no validation.
B) A manual A/B test via a separate load balancer configured independently.
C) Replacing the old version entirely at cutover.
D) Traffic splitting across multiple model versions behind a single deployment endpoint.
Show solution
Correct answers: D – Explanation:
Traffic splitting in model deployments routes a percentage of requests to a new model version while most traffic stays on the stable version — the canary safety pattern. Option C has no safety margin. Option B bypasses the deployment model. Option A has no validation. Source: Check Source
Question #5 - Model Deployment and MLOps
A deployed model emits prediction logs and latency metrics. The SRE team wants to alert when p99 latency exceeds 500 ms for 5 minutes, with automatic notification to a Slack channel.
Which OCI services chain together for this MLOps alerting?
A) Model deployment logs to Logging, metrics to Monitoring with an alarm, and Notifications to Slack.
B) Ignoring latency and relying on user complaints.
C) A spreadsheet updated hourly.
D) A manual review of latency in the console each morning.
Show solution
Correct answers: A – Explanation:
Logs -> Monitoring -> Notifications is the native OCI chain for MLOps alerting on model deployments. Option D and Option C are lagging. Option B is reactive. Source: Check Source
Question #6 - Data Access, Governance, and Integration
A notebook session needs to read a large Parquet file from Object Storage without hard-coding credentials or running as a user account.
Which authentication pattern fits?
A) Disabling authentication on the bucket.
B) Resource principals on the notebook session authorized by an IAM policy for the bucket.
C) Hard-coded API keys in the notebook.
D) The scientist’s personal access key pasted into the code.
Show solution
Correct answers: B – Explanation:
Resource principals give notebook sessions their own identity; IAM policies scope access to specific buckets — no credentials in the notebook. Option C and Option D are security violations. Option A exposes data. Source: Check Source
Question #7 - Model Development and ADS
A new data scientist wants to publish a model to a central location where the team can version it, see metadata (framework, training data reference, metrics), and deploy from there.
Which OCI Data Science feature supports this model catalog?
A) Emailing model files to the team each release.
B) A shared Excel file tracking models by hand.
C) Model Catalog (registry) with version, metadata, and deploy actions.
D) A local folder on a laptop.
Show solution
Correct answers: C – Explanation:
The Model Catalog stores models with versions and metadata, and supports deployment directly from a registry entry — the designed hub. Options A, B, and D are outside the system and lack governance. Source: Check Source
Question #8 - Jobs and Pipelines
A team has a multi-step ML workflow: ingest from Object Storage, feature-engineer, train, evaluate, conditionally publish. They want to orchestrate these steps with dependencies and partial re-runs.
Which OCI Data Science feature fits a multi-step ML workflow with dependencies and partial re-runs?
A) A PDF workflow diagram attached to an email.
B) Five unrelated cron jobs on different VMs.
C) A single notebook executed start-to-finish manually.
D) ML Pipelines that chain steps with dependencies and support partial re-execution.
Show solution
Correct answers: D – Explanation:
OCI Data Science Pipelines orchestrate multi-step ML workflows with dependencies and re-execution — the designed orchestration. Option C is monolithic. Option B is disjoint. Option A is not automation. Source: Check Source
Question #9 - Data Access, Governance, and Integration
A data platform team wants notebooks to read directly from an Autonomous Database without copying data to Object Storage first.
Which pattern supports direct Autonomous Database access from a notebook?
A) The Autonomous Database ADS integration using database credentials from OCI Vault or wallet-based auth.
B) Accessing Autonomous Database only through BI Publisher.
C) A manual copy of the full database to a CSV for each analysis.
D) Printing query results from a Toad session.
Show solution
Correct answers: A – Explanation:
ADS provides direct Autonomous Database connectivity with wallet-based auth or credentials in OCI Vault, reading data in place. Option C is slow and stale. Option D is outside the notebook. Option B is reporting only. Source: Check Source
Question #10 - Model Deployment and MLOps
A governance lead wants each deployed model to log every prediction request and response so the team can audit and detect concept drift later.
Which model deployment configuration captures request and response logging?
A) A daily ASCII dump to a user laptop.
B) A spreadsheet manually updated after each prediction.
C) Enable model deployment access and predict logs writing to OCI Logging.
D) No logging; rely on the model’s internal memory.
Show solution
Correct answers: C – Explanation:
OCI Data Science model deployments can emit access and predict logs to OCI Logging — the designed audit trail for drift analysis. Options A, B, and D are either not logging or not retained. Source: Check Source
Get 700+ more questions with source-linked explanations
Every answer traces to the exact Oracle documentation page — so you learn from the source, not just memorize answers.
Exam mode & learn mode · Score by objective · Updated April 19, 2026
Learn more...
What the 1Z0 1110 OCI Data Science Pro exam measures
- Model development and ADS (35%) — launch notebook sessions, manage conda environments, feature-engineer with Accelerated Data Science, run AutoML, and publish to the model registry.
- Jobs and pipelines (20%) — orchestrate repeatable training runs and multi-step workflows with OCI Data Science Jobs and Pipelines.
- Model deployment and MLOps (25%) — deploy models, emit logs and metrics, and split traffic to canary new versions safely.
- Data access, governance, and integration (20%) — connect notebooks to Object Storage and Autonomous Database, register assets in Data Catalog, and authorize service calls via resource principals.
How to prepare for this exam
- Review the official 1Z0-1110 exam page to confirm the current objectives.
- Complete the Oracle University OCI Data Science Professional learning path on MyLearn.
- In an OCI tenancy, launch a notebook session, train an ADS-assisted AutoML model on a public dataset, publish to the model registry, and deploy with a 10% canary traffic split.
- Apply the skills at work: move a Jupyter pipeline into OCI Data Science, wire a Job to run weekly, or register a feature dataset in Data Catalog.
- Master one objective at a time, starting with model development and the ADS library since it carries the most weight.
- Run PowerKram learn mode to see feedback after every question with sourced links back to Oracle documentation.
- Finish with PowerKram exam mode across all objectives until you clear the threshold three times in a row.
Career paths and salary outlook
OCI Data Science skills support machine-learning engineering careers:
- Data Scientist — $110,000–$160,000 (U.S. Bureau of Labor Statistics).
- Machine Learning Engineer — $130,000–$185,000 (Levels.fyi).
- MLOps Engineer — $120,000–$170,000 (Glassdoor).
Official resources
Work through the OCI Data Science Professional Learning Path on Oracle MyLearn. Supplement with the OCI Data Science documentation and the Accelerated Data Science library reference.