Yieldr AI"s homepage
Yieldr AI's Portfolio Tracker with Sample Data
Options Recommendation Engine
Yieldr AI's Comprehensive Dashboard
Options Recommendation
Yieldr's Menu

Yieldr AI: Turn Invested Stocks into Steady Income

Inspiration for the Idea

I’ve been actively investing for 6 + years, since I was in 8th grade, but I mostly invested and let my stock grow, using low risk options strategies on slow moving stocks in my portfolio like Pfizer($PFE), which helped lower my average cost basis per stock while the stock kept gaining value.

When I started talking with my dad's friends and other investors, I realized a lot of retail investors don't adopt the same strategy with their stocks. *Why?

Time - It takes a lot of time to screen the various stocks they own, especially if they work a long job and have other responsibilities.
Knowledge - A lot of investors group safer strategies like covered calls in the same boat as riskier options, often relating them to the wild type of options trades seen online.
Risk- Investors don't see the need to take the risk given the above for these gains

I always had this idea in my head, but didn't put much thought to it since I was a finance major that had no coding experience or technical knowledge. Eventually though, I discovered Bolt and began toying with it, and I realized I could actually build this with AI. When I saw the advertisement in my email for the Bolt hackathon, I took the opportunity to finally automate what I was already doing on a weekly basis, and sparked my idea for Yieldr AI

What We Learned/ Key Takeaways while Building with AI

1. AI is there in terms of product design : Within 1-2 prompts of Bolt, we were able to one-shot an extremely competitive prototype UI wise, especially with placement and copy.

2. It is better to build with AI rather than having AI build for you : What we noticed with AI code generation is that if you leave room for AI to make decisions itself, it usually deviates from your plan and causes issues. The best practice with bolt is to write what you would like to do, have it build out a plan of what it wants to implement, then have it implement it piece by piece.

3. Make minor changes directly in the editor : If you want to make minor changes like text or sizing, its easier to directly edit it in the bolt.new editor rather than prompting the AI, both for tokens usage and preserving the rest of our code.

How We Built It (Stack)

Front-End
- TypeScript / JS
- Deployed on Netlify via custom domain provided by IONOS for the Bolt Hackathon
Back-End
- Supabase
- E-Trade API for real-time stock prices, option premiums, strike price, delta, etc.
- AI recommendations via the Claude API, with a custom model and curated system prompt