Stepwise reasoning with DSPy

This example shows a simple reasoned response pattern with DSPy.

It keeps the model local and deterministic, but still returns a reasoning field and a final answer field so you can see the shape of a stepwise DSPy output.

Requirements

• Python 3.10+

Steps for running this example in the shell

1. Install dependencies:

python -m venv .venv

# macOS/Linux:
source .venv/bin/activate

# Windows Command Prompt:
.venv\Scripts\activate.bat

# Windows PowerShell:
.venv\Scripts\Activate.ps1

pip install -r requirements.txt

2. Run the script:

python stepwise_reasoning.py

What it demonstrates

• A DSPy signature with a reasoning field and a final answer
• A local LM stub that emits repeatable JSON
• A small offline demo with no external dependencies

Output

Question 1: How should we handle the product launch this week?
Reasoning: A launch request needs a short plan and a clear final status.
Answer: Summarize the release steps, confirm the rollout owner, and share the final checklist.

Question 2: What should we say about the duplicate billing report?
Reasoning: Billing questions should confirm the charge and keep the reply precise.
Answer: Verify the charge, explain the refund status, and point the user to billing support