A clear route through the work.
Twenty weeks of connected modules, followed by a four-week project that brings the whole delivery cycle together.
Choose a sustainable pace
The core assumes working knowledge of Python, SQL, Git, and HTTP. These are planning estimates, not a promise of job readiness. Most time belongs to implementation, review, and portfolio work.
| Path | Duration | Weekly effort | Planned total |
|---|---|---|---|
| Optional foundations | 4 weeks | 16 hours | 64 hours |
| Core + capstone | 24 weeks | 16 hours | 384 hours |
| Accelerated core | 12 weeks | 32 hours | 384 hours |
A typical week allocates four hours to concepts, eight to practical work, two to delivery artifacts, and two to assessment and reflection.
The 24-week sequence
Understand the work and platform
Client discovery, production Python, SQL, and cloud/Databricks foundations. Deliver a brief, API contract, data model, and access matrix.
Build reliable data and retrieval
Spark/Delta, orchestration, LLM applications, and RAG. Deliver a repeatable pipeline and an evidence-based retrieval baseline.
Establish the trust boundary
Governed search, bounded agents, MLflow evaluation, and security. Deliver access tests, approval behavior, and a reproducible evaluation report.
Integrate and justify tradeoffs
CI/CD, enterprise integrations, domain prototyping, and quality/latency/cost. Deliver a release path and an evidence-backed architecture decision.
Prepare to operate and hand off
Deployment, incidents, client delivery, and portfolio practice. Deliver a runbook, rollback rehearsal, and technical/executive narratives.
Complete the integrated capstone
Discover, implement, harden, and defend a governed support copilot. Submit the complete evidence bundle for review.
Progress is evidence
Use the lesson checks as a personal study aid. Practical competence comes from producing and reviewing the artifacts. A course completion check is not an employer credential or vendor certification.
Start with the diagnostic