1 minute read

Every year, my wife and I go through the same process - reviewing our investment portfolio, gathering data, checking allocations, and comparing fund performance. It’s tedious, manual, and spreadsheet-heavy. This year, we decided to automate it a bit.


In a single day, I built a tool that takes a screenshot or CSV of a portfolio, extracts structured data, runs an analysis, and provides insights. It helps answer key questions:

  • What funds do I hold, and how do they compare to benchmarks?
  • Are my allocations in line with my targets?
  • What are my total fees and expense ratios?
  • Are there better fund options with lower fees and higher ratings?






The tool uses (and was built with) OpenAI, Repo Prompt, and o1 pro, automating most of the coding process. Instead of manually writing every function, I used Repo Prompt and o1 pro to generate much of the code, making it easier to iterate and expand.

One of the most interesting parts of this project was adding a new feature using only voice commands — I described the change I wanted, and with the help of Repo Prompt, o1pro wrote the code. In the video, I walk through this step-by-step, showing how I used AI to build and modify the tool in real time.

This kind of workflow makes it possible to build complex applications quickly without needing to write everything from scratch. If you’re interested in how AI tools like this can speed up development, check out the video above!

Let me know what you think — or if you’re working on something similar, I’d love to hear about it.

Updated: