What problem does it solve?
Developing complex AI applications often requires manual prompt engineering, brittle pipelines, and repetitive code, making it hard to scale and maintain reliable language model programs.
Core Features & Use Cases
- Declarative signatures: Define inputs and outputs with type‑safe signatures instead of ad‑hoc prompts.
- Automatic prompt optimization: Use built‑in optimizers (BootstrapFewShot, MIPRO) to generate effective few‑shot examples.
- Modular pipelines: Compose modules like Predict, ChainOfThought, ReAct, and ProgramOfThought into reusable RAG or agent systems.
- Use Cases: Build question‑answering bots, retrieval‑augmented generation pipelines, self‑consistent classifiers, or multi‑step agents without hand‑crafted prompts.
Quick Start
Ask the dspy skill to create a chain‑of‑thought module that answers “What is the capital of France?”.