A few days ago, I posted asking whether our company supports AI tools.
I wanted to use Claude Code, but since the company doesn't provide support for it, I ended up paying $100 a month for Claude Code Max out of my own pocket.
And thinking I should do something with the $100 I spend every month, I recently tried applying it to stock investment, which I've been very interested in.
I became interested in domestic stocks when the Lee Jae-myung administration came in and started investing, but since I haven't been investing for very long, when I look at terms like PER and ROE or financial statement figures, I can't really tell if something is good or bad. So I thought it would be better to ask an AI instead, and I created an investment analysis web service using Claude Code with a colleague from work who's also interested in investment, working on it in our free time after work.
When you ask about a stock, an AI Agent compiles a comprehensive report based on charts, financial statements, and analyst opinions, and also explains unfamiliar terms and figures. You can think of it as that kind of functionality.
In my case, I bought and continue to hold a KOSPI 200 ETF with a relatively large amount last year, and for individual stocks, I occasionally ask through the feature I created, and when positive aspects come out highly, I buy with a small amount.
If you really look at the screenshot, you'll see that without Claude Code it would have taken a very long time to build, but it turned out to be made much faster than expected.
However, Claude Code isn't perfect. It actually codes really well, but sometimes it refactors code that shouldn't be touched, which often causes features that were working to break.
Below is a sample report of a stock analysis I did today.
https://nubble-storage.nu-us.com/noa/shared-pages/14d952f5.html
I'm curious if there are others who do investment analysis with AI and how you're using it, and what LLM model you're using.
For reference, I'm using Claude 4.6 Opus.
