AI / ML Path

From structured LLM outputs to a complete 12-month DSPy mastery curriculum.

Prerequisites: Jinja2 and Pydantic from the Getting Started path.

graph TD
    instructor["instructor"] --> dspy["DSPy Series<br/><small>12 sessions</small>"]
    instructor --> rlm["RLM Project"]
    instructor --> mbot["marketing-bot"]
    dspy --> sline["Sline Case Study"]

    style instructor fill:#e1bee7,stroke:#6a1b9a
    style dspy fill:#e1bee7,stroke:#6a1b9a
    style rlm fill:#e1bee7,stroke:#6a1b9a
    style mbot fill:#e1bee7,stroke:#6a1b9a
    style sline fill:#e1bee7,stroke:#6a1b9a

    click instructor "../wiki/lightning-talks/instructor/"
    click dspy "../wiki/series/dspy-mastery/"
    click rlm "../wiki/projects/rlm/"
    click mbot "../wiki/projects/marketing-bot/"

The Sequence

1. Instructor — Structured LLM outputs with Pydantic models. Uses OpenAI API + Instructor library to generate validated, typed responses from language models.
2. DSPy Mastery Series — 12 sessions (June 2025 – May 2026) covering the DSPy framework: LM setup, data collection, signatures, adapters, modules, metrics, optimization, assertions, and production tracking.
3. RLM Project — Recursive Language Models in 77 lines. Literate programming approach where slides are the source of truth for code.
4. marketing-bot — Clean architecture applied to AI systems: Deming cycle (Plan-Do-Check-Adjust), dependency injection, abstract base classes, structured output.

Where to Go Next

• DSPy Session 6 (Metrics) connects to → Testing & Quality
• marketing-bot architecture patterns apply broadly to any Python project