Most candidates pass with 80 to 160 focused hours depending on ML background. The best use of time is to study by lifecycle stage and object choice, not by memorizing disconnected feature lists.
How to use this plan well
Each study block should do four things:
- classify the question by lifecycle stage
- decide which OCI object or control belongs there
- do a short scenario set
- write down whether the miss was evaluation, deployment, monitoring, or governance
flowchart LR
Classify["Classify stage"] --> Object["Choose object or control"]
Object --> Drill["Do short scenario set"]
Drill --> Review["Review why misses happened"]
Review --> Classify
How long should you study?
| Your time |
Recommended timeline |
Good fit |
| 18 to 22 hrs/week |
30 days |
intensive path with recent ML background |
| 10 to 14 hrs/week |
60 days |
balanced path for most candidates |
| 6 to 9 hrs/week |
90 days |
part-time path with slower reinforcement |
30-day intensive plan
| Week |
Focus |
Output |
| 1 |
OCI Data Science objects, projects, notebook sessions, jobs, and data handling |
object-choice notes and short drills |
| 2 |
training, evaluation discipline, leakage, metric fit, and artifact handling |
evaluation tie-break sheet |
| 3 |
deployments, monitoring, rollback, and latency or drift response |
weak-lane notes and mixed sets |
| 4 |
governance, security, repeatability, and final readiness |
mixed sets and compression |
60-day balanced plan
| Phase |
Weeks |
Focus |
| 1 |
1 to 2 |
terminology cleanup and lifecycle-stage classification |
| 2 |
3 to 4 |
object choice, data prep, and repeatable execution |
| 3 |
5 to 6 |
training, validation, leakage, and metric fit |
| 4 |
7 |
deployment, rollback, and monitoring |
| 5 |
8 |
governance, access, and operational traceability |
| 6 |
9 to 10 |
weak-lane repair and final mixed review |
90-day part-time plan
| Month |
Focus |
Goal |
| 1 |
objects, lifecycle vocabulary, and data handling |
stop losing points to stage confusion |
| 2 |
evaluation, model artifacts, and deployment boundaries |
build stronger lifecycle judgment |
| 3 |
monitoring, governance, and exam-style tie-breaks |
finish with mixed-set confidence |
If misses cluster here, do this next
| Miss pattern |
Weak lane |
Fix next |
| you keep choosing the wrong project, notebook, job, model, or deployment |
object choice |
review object boundaries and use cases |
| you trust good metrics too quickly |
evaluation |
review leakage, split quality, and metric fit |
| you ignore rollback, monitoring, or latency |
delivery and ops |
review deployment safety and observability |
| you mix model-quality problems with platform-health problems |
lifecycle separation |
review stage boundaries and control responsibilities |
What strong prep usually does
- classifies the stage first, then picks the OCI object or control
- keeps a short confusion list for job vs deployment, training vs inference, and artifact vs dataset
- writes down why the winning answer is safer or more repeatable instead of just memorizing it
- uses Oracle docs to settle disagreements, then comes back here for compression
Final 72 hours
| Keep doing |
Stop doing |
| rereading the cheat sheet and glossary |
opening unrelated new ML services |
| reviewing weak-lane misses |
treating every question like a theory question |
| checking official docs for disputed boundaries |
building a large new pipeline or deployment late |
| practicing lifecycle and object classification |
trusting unsupported community notes over Oracle docs |
Route yourself well
- lifecycle traps and object tie-breaks: Cheat Sheet
- high-confusion data-science terms: Glossary
- last-week questions: FAQ
- official Oracle and OCI source routing: Resources