Skip to content

DSPy Mastery Series
Duration 12 sessions (June 2025 – May 2026)
Status Ongoing
Prereqs Instructor, Pydantic
Repo pysprings/dspy-sequence

Overview

A structured 12-month curriculum covering DSPy, the Python framework for systematic AI development. Each session builds on the last, progressing from problem framing through production-ready AI systems.

Session Calendar

# Title Focus
0 AI Development the Python Way Problem framing: the black box challenge, prompt engineering pain
1 LM Setup The Language Model class as gateway to AI services
2 Data Collection Building training data foundations
3 Signatures Declaring what you want — interface design for AI
4 Adapters Translation layers between specs and execution
5 Basic Modules Working AI programs
6 Metrics Measuring success
7 Optimization The DSPy superpower — automatic prompt optimization
8 Advanced Modules Sophisticated patterns
9 Assertions Building reliable AI systems
10 Trackers Production observability
11 Retrospective Mastery achieved

Case Study: Sline

The series includes a hands-on case study: Sline, a shell command assistant that converts natural language to bash commands. It demonstrates the end-to-end DSPy workflow: signature design, training data creation, custom adapters, metrics, and optimization.

Getting Started

  1. Clone: git clone https://github.com/pysprings/dspy-sequence
  2. Each session has its own directory with presentation.md and code/
  3. Start with Session 0 for context, then follow sequentially

Where to Go Next