DSPy: The framework for programming—not prompting—language models
Python
Update index.md
Omar Khattab4e59b13b38c2
diff --git a/docs/docs/index.md b/docs/docs/index.mdindex 2ba0f63d1..273e7dac5 100644--- a/docs/docs/index.md+++ b/docs/docs/index.md@@ -14,7 +14,7 @@ hide:DSPy is a declarative framework for building modular AI software. It allows you to **iterate fast on modular code**, not brittle strings, and offers algorithms that **compile modular AI programs into effective prompts and weights** for your models, whether you're building simple classifiers, sophisticated RAG pipelines, or Agent loops.-DSPy stands for Declarative Self-improving Python. Conventionally, the most important part of your ML system is a sequence of hand-stitched training jobs or long string prompts. Instead, DSPy lets you **define AI software by declaring natural-language modules** and _generically composing them_ with different models, inference strategies, or learning algorithms.+DSPy stands for Declarative Self-improving Python. Instead of hand-stitched training jobs or long prompts, it lets you **define AI software by declaring natural-language modules** and _generically composing them_ as first-class constructs with different models, inference strategies, or learning algorithms.This abstraction makes your AI software more reliable, maintainable, and portable as new models and learning techniques emerge. It's also just rather elegant. This [lecture](https://www.youtube.com/watch?v=JEMYuzrKLUw) is a good conceptual introduction. Meet the community, seek help, or start contributing via our [GitHub repo](https://github.com/stanfordnlp/dspy) and [Discord server](https://discord.gg/XCGy2WDCQB).