Covered Topics

Part 1: Foundations - How DSPy Changes Everything

• Understanding the DSPy Paradigm: Why shifting from prompt hacking to high-level programming matters for reliability
• DSPy Internals: How the framework compiles declarative code into optimized prompts under the hood
• Modules and Signatures: Building your first reusable AI components with clear input/output contracts
• Development Setup: Configuring local models (Ollama, SGLang) and cloud providers for immediate productivity

Part 2: Building Intelligent Applications

• Chain-of-Thought and Reasoning: Implementing complex reasoning strategies that actually work
• RAG Systems That Scale: Building retrieval-augmented generation with Weaviate, Milvus, and ColBERTv2
• Agentic AI and Tool Use: Creating AI agents that plan, use tools, and interact with external systems via native async support for dspy.Tool
• Multi-Stage Pipelines: Composing modules for document processing, analysis, and generation workflows

Part 3: Optimization - The Secret Sauce

• Automatic Prompt Engineering: How BootstrapFewShot and MIPROv2 eliminate manual prompt tuning
• Self-Improving Pipelines: Setting up feedback loops that continuously enhance performance
• Fine-Tuning Integration: When and how to use BootstrapFinetune for model weight optimization
• Custom Metrics: Designing evaluation criteria that drive meaningful improvements

Part 4: Production-Grade Patterns

• Structured Outputs: Enforcing JSON schemas with TypedPredictors and Pydantic validation
• Error Recovery: Building resilient systems with automatic retries and schema-aware feedback