Step by Step: How Claude Built the RTM Trading Bot

In the previous video, we saw Claude build a trading robot, but it wasn't the coding that surprised me. It was the conversation. The number of times Claude spotted things that I missed, challenging my assumptions, or completely changed the direction we were heading. So, this video isn't only about the trading bot. Well, it is, but it's

also about the journey that built it. All the dead ends, all the wrong turns, the ideas that survived, and the ones that didn't. By the end, we walk away with this beautiful river to mean trading bot, but the real story is how we got there. So, with no more shenanigans, let's start at the

beginning. Now, the idea I had is very simple. What we have is the average price in the market in the form of EMA 200. It could be EMA 400, it could be EMA 100, but an average. And if the price moves too far away from that average, then we think that there is a high likelihood that the price is

going to come back. Now, what is too much? Is that too much? Is this too much? Is that too much? So, how I thought to do it was that if the price has moved away 3% or more, and the RSI has gone into the oversold territory, that's when I would like to enter the

market. Now, the more we would want the price to be stretched away from the EMA, the higher the likelihood of it turning back. But, on the other hand, these kinds of price level might not happen that frequently. So, it's a trade-off between having more trades versus having safer trades. So, is two enough? Is two and a half

enough? Is three enough? And that's when I decided to leave it as a user input, so we can run different optimization models and see which value holds the best trade-off. But, along the way, as I was speaking with Claude, I realized that this 3% away from the EMA is not going to hold across time frames. On a