Days 1-30
Data quality, perception-data readiness, Python/SQL workflow design, and API fundamentals.
Learn production AI engineering by building.
This is a self-paced technical course following an AI data and MLOps delivery path, with explanations, system diagrams, worked examples, labs, quizzes, interview prompts, and evidence gates that grow into a portfolio capstone.
Course
Lessons
Specialization
Health AI Engineering Lane
Portfolio Project
National Health AI Readiness and Triage Copilot
A synthetic-data-only capstone that demonstrates data quality checks, model evaluation, API serving, RAG, monitoring, documentation, and responsible deployment patterns for health-system AI.
Read Project SpecExecution
90-Day Learning Plan
Days 31-60
Kubernetes jobs, Databricks, manifests, retries, artifact gates, CI/CD, and observability.
Days 61-90
Customer-facing delivery, health AI, synthetic-data evaluation, responsible AI, and capstone documentation.
Primary-source desk ยท checked 23 September 2026
Current engineering references
Reliable batch work
Kubernetes Jobs documents completion, retries, backoff limits, parallel work, and failure policy. Design tasks to tolerate reruns; do not assume a controller guarantees exactly-once execution.
Orchestration and observability
Databricks Lakeflow Jobs covers task dependencies, triggers, run history, notifications, and workflow monitoring. Prefect retries is a practical companion for bounded retry policies.
Responsible evaluation
NIST AI RMF and its AI Resource Center provide voluntary guidance for governing, mapping, measuring, and managing AI risk. Apply it to the synthetic health-AI capstone; do not use real patient data.