Decision Intelligence and AI: Systematically Improving Decision Quality with Generative AI

This open-source project 'Decision Intelligence with AI' is an interactive course combining decision theory and generative AI, accompanying an upcoming book. It aims to help individuals and organizations systematically improve decision quality. The project provides a complete workflow from decision frameworks to AI implementation, including Jupyter Notebook modules and C#/.NET code examples. It supports multi-path learning design, catering to users with diverse backgrounds from non-technical to AI engineers.

Decision-making ability is one of the most valued business skills by executives, but making effective decisions is not easy. This project explores the intersection of decision theory and generative AI. As an open-source interactive course accompanying a book, it adopts a multi-path learning design. Whether readers are from technical or non-technical backgrounds, they can find a learning method suitable for themselves, with content progressing from basic concepts to enterprise-level applications.

The project revolves around the decision intelligence framework, broken down into six interconnected stages: 1. Introduction to decision frameworks (e.g., Eisenhower Matrix combined with AI); 2. Decision framing (reconstructing options, Six Thinking Hats technique); 3. Intelligence gathering (historical cases, AI-assisted integration, collective wisdom); 4. Decision execution (intuition, rules, quantitative methods like Monte Carlo simulation); 5. Decision communication (Minto Pyramid Framework, AI to reduce bias); 6. Enterprise-level decision intelligence (building decision support/management systems).

The technical implementation is based on the C#/.NET ecosystem, using Semantic Kernel and AI Extensions frameworks. It includes modules such as simple decision prompts, Chat Completions API, reusable prompts and native functions, and private AI decision workflows. There are three learning paths: Reading (non-technical, reading pre-run content in the browser), Technical Basics (running code via Jupyter Notebook), and AI Engineer (deeply modifying code, porting the project).

The project includes multiple examples of practical decision scenarios that can be directly applied or adapted: School choice decision between community college vs. traditional university, total cost analysis of car purchase, sales team performance optimization, resource allocation trade-off between exploration and exploitation, intuitive decision-making in time-sensitive situations, etc.

The project is accompanied by the upcoming book 'Decision Intelligence with AI'. The book focuses on theory and advanced cases (suitable for non-technical users), while the project emphasizes code practice (suitable for hands-on learners). Currently, the project is in active development, with a Beta version expected to be released in August 2026. More than 10 modules are already available, and the remaining modules are being migrated to Microsoft Extensions for AI and Agent Framework.

The core value of this project lies in providing a systematic decision-making methodology, connecting decision science research results with generative AI capabilities, and offering actionable decision improvement solutions for individuals and organizations. For developers: It demonstrates how to embed AI into decision workflows; For decision-makers: It provides a structured method to improve decision quality; For researchers: It connects decision theory with AI practice. In the AI era, 'making better decisions with AI' is a core competency, and this project is a high-quality resource.