⚠️ This column shares the process of creating prompts for free educational purposes and does not constitute investment, legal, or financial advice.** AI execution results may contain errors, so independent verification is essential. All responsibility for their use lies solely with the user. They cannot be directly used for business decisions. [See all essential checklists below the 30-week timeline before use[]](https://www.notion.so/30-29f84ddcec3c80a1a87eda9883d81453?pvs=21)
I majored in economics at a late age, aiming to become a full-time investor. Before graduation, I read and analyzed nearly 300 investment-related books to find the best investment strategy. For roughly 50 months after graduation, I traded in international futures markets, including the S&P 500, crude oil, and U.S. Treasury bonds, making thousands of adjustments to my strategy.
Every morning, before the market opened, I checked international news and analyzed risks. A single tariff policy or central bank announcement could shake the market, and I had to adjust my positions accordingly. However, this entire process was manual.
I checked the news every day, verified my investment strategy, adjusted it, and measured my returns. I spent an average of five hours a day on risk management, but I was constantly anxious. "Have I checked all the important news?" This question kept running through my head. "What if I missed something?" I never felt a moment of peace.
There was always a demand for automation. However, I wasn't a developer and had no knowledge of code. I attempted automation several times, but I couldn't solve various problems, such as the security of my trading strategy and the cost of development, and continued to do so manually.
Then, around June 2025, witnessing the advancement of AI technology, I began my journey to automation in earnest. Claude's Model Context Protocol (MCP) was the starting point.
The era had finally arrived where AI could go beyond simply answering questions and automatically handle complex tasks. Seeing AI execute requests in natural language instead of writing complex and difficult code, I felt I could finally live with peace of mind.
First, I wanted to free myself from the 23-hour grind of overseas futures trading by building an automated trading system. Second, I wanted to develop a program to verify profitability. Having read hundreds of books and tested numerous strategies, and even revised them more than I tested during actual trading, I longed to see AI calculate profitability just like I did.
However, I wanted to reduce the time spent on information gathering, so I started by automating my daily analysis of international political trends. For several weeks, I designed prompts that would allow the AI to analyze and verify my insights, along with examples, just like I did. This process gradually evolved, with file size segmentation and seven hub systems, operating almost like a single program.
Over time, my insights grew stronger, my information selection ability improved, and when I delivered articles, the AI analyzed them, extracted the necessary keywords, and verified the information. Ultimately, I successfully utilized AI to analyze international political trends. Processing all this in under an hour significantly increased efficiency, and I realized I needed to leverage AI more fully.