Debugging Complex Logical Workflows with Chain-of-Thought

AI Skill Booster

Important note: This training may demonstrate AI tools that are not approved for use with Stanford data. Inclusion in this session does not imply institutional approval. Participants should refrain from entering Stanford data into unapproved tools. An up-to-date list of approved and reviewed tools is available on the GenAI Evaluation Matrix page.

Code	Date	Delivery	Cost
ITSB-1013	Tue Jun 2, 1:00 pm to 2:30 pm	Live Online : 1 session	$150

Before each live online session, Tech Training will provide a Zoom link for live online classes, along with any required class materials.

Practice using Chain-of-Thought prompting to troubleshoot broken logic, code, and workflows. Explore how to guide AI through step-by-step reasoning to surface errors and identify solutions.

Program Description

When standard AI prompts fail on complex, multi-step problems, Chain-of-Thought prompting can help. This technical session explores how to instruct AI to show its reasoning process, making it possible to identify exactly where logic breaks down. Participants practice applying CoT techniques to debugging scenarios involving code, workflows, and analytical problems.

Learning Objectives

Learners will have the opportunity to:
1. Explore how Chain-of-Thought prompting surfaces hidden errors in complex workflows
2. Practice constructing prompts that force AI to reason through problems step by step
3. Work with CoT techniques to troubleshoot broken code and identify root causes
4. Experiment with AI-assisted debugging across logical and analytical scenarios
5. Apply CoT prompting to a real problem or workflow from your own work

Topic Outline

Topics:
- Why standard prompts break on complex logic problems
- Chain-of-Thought prompting for debugging and analysis
- Prompt structures for step-by-step reasoning
- Applying CoT to code, workflows, and data problems
- Verifying and validating AI reasoning outputs

Tools used: Stanford AI Playground and free versions of other AI models (ChatGPT, Gemini, or Claude)

Custom training workshops are available for this program

Technology training sessions structured around individual or group learning objectives. Learn more about custom training

Special Group Rates

For groups of 5 or more within the same team or department, special rates are available. Please contact techtraining@stanford.edu for more details.

University IT Technology Training sessions are available to a wide range of participants, including Stanford University staff, faculty, students, and employees of Stanford Hospitals & Clinics, such as Stanford Health Care, Stanford Health Care Tri-Valley, Stanford Medicine Partners, and Stanford Medicine Children's Health.

Additionally, some of these programs are open to interested individuals not affiliated with Stanford, allowing for broader community engagement and learning opportunities.