Painless AI Lab – A Hong Kong Practitioner’s Real-World Journey (Part 7)

After years of chasing market tips and watching friends lose money to emotional decisions, I finally reached a breaking point. I was tired of guessing. I wanted a clear, objective way to analyze companies — looking at real numbers, business risks, and actual potential instead of the usual hype. That desire led me to build “Find the Right Stock” — my own web application.

This post isn’t a technical manual. It is the story of how I designed my own system to solve a problem that has plagued my investment strategy for years.

The Philosophy: Why I Stopped Guessing

In Hong Kong’s fast-moving market, FOMO is the enemy of wealth. I found myself making the same mistakes as everyone else: buying on “hot tips” and selling in panic. I realized that my learning was scattered because I lacked a clear, personal goal.

So, I decided to create my own “Objective Financial Filter.” By building this system, I regained control of my process. I no longer rely on gut feeling or market noise. Instead, I now have a repeatable methodology to:

• Check & Verify real-time financial data.
• Stress-test a company’s business model.
• Analyze macro trends without the distraction of daily gossip.

My System Design

I wanted a tool that was clean, professional, and entirely my own. I focused on four pillars:

• The Intelligence Engine: A custom-configured logic model that acts as my dedicated research analyst.
• The Data Foundation: A framework that pulls live stock prices and verified financial statements to keep my analysis grounded in fact.
• The Analysis Logic: The most critical part. I spent weeks iterating on a custom prompt—the “Financial Logic Brief”—that forces the system to calculate fair value, identify hidden risks, and suggest a target price based on my specific criteria.
• The Delivery: I designed this as a lightweight web app, allowing me to generate professional reports on the go.

The “Vibe Coding” Workflow

I am not a professional engineer, but I am a practitioner who believes in leveraging modern tools to execute a vision. I used a “Vibe Coding” approach—a process where I direct the logic and the AI handles the translation into functional code.

My Methodology:

• Intelligence: Ollama + Local Models (Gemma 4, Qwen, etc.)
• Logic Builder: Google AI Studio
• Infrastructure: GitHub and Vercel
• Integration: WordPress (Hosting)

My Process:

1. Define the Vision: I mapped out exactly what I needed the tool to output for a given stock ticker.
2. UI Design: I kept the interface minimal, focusing on readability and professional utility.
3. Prompt Engineering: This was my personal work. I spent hours refining the system instructions to ensure the tool didn’t just give me “generic” advice, but calculated metrics based on my proprietary logic.
4. Backend & Deployment: I configured the data streams to ensure accuracy and deployed the final product to my site.

The Most Important Lesson

The takeaway from this project wasn’t just the code; it was the realization that when you build a system to solve a real problem in your own life, the learning becomes permanent.

Every time I use this tool, I am not just getting a report—I am training my own judgment. I have moved from “following the market” to “interrogating the market.”

Question for you: Have you ever built a personal tool to solve a problem in your work or investments?

What is your strategy for staying objective when the market gets emotional?

Share your thoughts below.

Explore my system here: www.ib-hk.com (projects)