How AI Has Transfored Our Trading Results

Sep 26, 2025

4 min read

A real game changer and the results are proving it

I’ve been developing trading strategies for many years, and for a long time, my results were respectable. I had what I would call “pretty good” software—systems that worked, strategies that were profitable, and processes that I had refined through years of experience. Yet, there was always an Achilles heel. No matter how much I improved a system, I could never quite eliminate that one weakness or blind spot. My results consistently fell short of the standards I had set for myself.

I wanted more.

The Turning Point: AI Enters the Picture

The introduction of Artificial Intelligence—particularly through tools like ChatGPT—has been nothing short of a gamechanger for my work. Rather than patching holes in existing systems, I decided to go back to square one and see if AI could help me uncover entirely new approaches.

So, I sat down with ChatGPT and did a massive “brain dump” of my trading knowledge, outlining everything I had learned over the years along with my aspirations for what I wanted to build. To my surprise, the AI responded with ideas and directions I had never considered before. Some concepts were subtle adjustments that refined my thinking; others were entirely new perspectives that broadened what I thought was possible in strategy design.

For the first time, I felt like I had a partner in problem-solving—a system that could process my experience, challenge my assumptions, and generate fresh, actionable insights.

Defining the Perfect Portfolio of Bots

From the start, I knew that one of my key requirements was to build a portfolio rather than rely on a single strategy. The portfolio had to contain no more than eight strategies, each with its own distinct personality: a mix of different currency pairs, different trading styles, and different market conditions.

But there were more rules:

The system needed to deliver higher returns than what I had achieved previously, without adding more risk.
The design had to be simple and reliable, avoiding unnecessary complexity.
It had to handle periods of extreme volatility with ease, not just in terms of trading performance but also in backend efficiency—keeping CPU loads low and reliability high.
Finally, it had to offer flexible risk/return profiles so that investors with different appetites could select the level of exposure that suited them best.

These weren’t easy requirements to balance. In fact, in the past, I struggled for years trying to reconcile them. But with the help of AI, I was able to filter through features worth keeping and discard those that added little value. Some of the best features I ended up implementing were things I had never even thought of before.

Building Through Collaboration

Once the framework was in place, I began developing and backtesting individual strategies. After each round of testing, I would feed the results back into ChatGPT, asking for insights, improvements, or new angles. This iterative process created a feedback loop between myself and the AI: I provided the raw experience and testing data, while it offered suggestions, refinements, and new approaches to explore.

What might have taken me years to slowly evolve on my own was accelerated dramatically. Within a matter of months, I had not only backtested but also live-tested strategies that exceeded my expectations in performance, reliability, and scalability.

Results Beyond Expectations

The outcome of this collaboration is a new family of trading strategies that are performing above and beyond what I thought possible. In real market conditions, they have shown stronger returns, greater stability, and more resilience to market shocks than any of my earlier systems.

AI didn’t just help me shorten the development cycle—it shaved years off the learning curve. What would once have been a drawn-out process of trial and error has now become an agile, adaptive system that continues to evolve as new data comes in.

Another key advantage is that the process doesn’t stop. As we continue to gather live results, those outcomes are fed back into the AI to identify potential weaknesses and improvements. It’s like having a constant research partner monitoring performance and suggesting upgrades in real time.

Not a Magic Bullet, But a Gamechanger

It’s important to acknowledge that AI isn’t a silver bullet. It doesn’t remove all the challenges of building successful trading strategies, and it doesn’t replace the need for human judgment, experience, and discipline. However, what it does provide is a dramatic acceleration in research and development. It offers a powerful template to work from and a creative partner to explore possibilities that I might not have seen on my own.

That’s why our latest release, the Institutional Edge series, represents such a leap forward. I deliberately chose not to brand it around AI—too often, “AI” is used as a buzzword that oversells and underdelivers. Instead, I wanted the performance to speak for itself. But make no mistake: the breakthroughs in this series are directly tied to what AI has made possible.