How I'm Using AI to Help Investors Follow Their Own Rules

Hey IH community! I'm William, and I've been building KeepRule (https://keeprule.com) for the past six months. I wanted to share what I've learned because the problem space is fascinating and the journey has been a rollercoaster.

The Problem I Kept Seeing

I've been investing for years, and I noticed something that kept happening — to me and to everyone I talked to about investing. We'd read Buffett's letters, study Munger's mental models, write down our investment rules... and then completely ignore them when the market got exciting or scary.

A friend of mine wrote "NEVER buy on FOMO" on a sticky note on his monitor. Three months later, he panic-bought a meme stock at the top. The sticky note was still there.

That's when it clicked: the problem isn't knowledge. Investors know what they should do. The problem is consistent execution under emotional pressure.

What KeepRule Does

KeepRule is an AI-powered investment principles platform. The core idea is simple:

1. Curated investment principles from masters like Buffett, Munger, and Graham — not just quotes, but structured rules with context
2. AI chat where you can discuss investment decisions with an AI trained on these principles. Think of it as a conversation with a Buffett-informed thinking partner
3. Scenario-based decision support — real investment situations where you can explore how the principles apply

It's not a stock screener or a robo-advisor. It's a tool for building and maintaining investment discipline.

What's Working

Some things that surprised me:

The "chat with Buffett" feature is sticky. I built it almost as an experiment, but users keep coming back to it. People use it to stress-test their investment theses before buying. One user told me he saved himself from a bad trade because the AI pointed out he was violating his own rule about debt-to-equity ratios.

Scenarios drive engagement. I added scenario-based content — real investment dilemmas where you apply principles to make a decision. These get way more engagement than static content. People want to practice, not just read.

SEO is working but slowly. I'm getting organic traffic for long-tail queries like "Buffett investment rules checklist" and "value investing decision framework." Not huge numbers yet, but the intent is high — these people are actively looking for exactly what KeepRule offers.

What's Not Working (Yet)

Conversion from free to paid is hard. The free content is valuable enough that people don't feel the urgency to upgrade. I'm experimenting with gating the AI chat features and advanced scenarios.

Social media is noisy. The investment content space is incredibly crowded. Standing out requires being genuinely useful, not just posting generic tips. I've had more luck with detailed, practical content than short-form posts.

B2B might be the real opportunity. I've had a few financial advisors reach out asking if they can use KeepRule with their clients. I hadn't considered this use case initially, but it makes sense — advisors need tools to help clients stick to their investment plans.

Numbers (Being Transparent)

• Launched the current version about 4 months ago
• Growing organically, primarily through content marketing and SEO
• Retention is encouraging — users who engage with the AI chat feature come back regularly
• Revenue is early stage but growing month-over-month

Tech Stack (For the Curious)

• Backend: Go (Go-Zero framework) — fast, clean, great for API services
• Frontend: Vue 3 + TypeScript with Nuxt for SSR
• AI: Claude API for the chat features
• Hosting: Alibaba Cloud (I'm based in Asia, latency matters)

What I'd Do Differently

1. Start with the AI chat, not the content library. The interactive features are what make KeepRule different. I spent too long building the static content first.
2. Talk to users earlier. My first 10 conversations with real users changed the product more than 3 months of building in isolation.
3. Don't underestimate SEO. For a niche product, long-tail SEO is incredibly effective. I wish I'd started writing content from day one.

What's Next

I'm focused on three things:

• Deepening the AI features (more personalized rule recommendations)
• Building out the scenario library (this is what keeps people coming back)
• Exploring the B2B angle with financial advisors

If you're curious, check it out at keeprule.com. Would love feedback from this community — you all have a great eye for product.

Happy to answer any questions about the build, the market, or the tech decisions!