Engineering Leadership Series - George Kailas
Biography
George Kailas is the CEO and Founder of Prospero.ai, where he leads the company’s mission to democratize access to institutional-grade financial insights for everyday investors. With over 14 years of experience in artificial intelligence and 23 years in professional investing, George brings a rare combination of deep technical expertise and lifelong market intuition to the role. George not only leads the company but also engineered the core platform himself
Enjoy the interview and give George a follow.
More Information
George Kailas
The Interview
- Tell us your story, how did you discover your interest in finance and markets?
I’ve always been curious about how systems work, and markets were the biggest and most dynamic system I could study in real time. In high school I taught myself accounting, landed an internship at a hedge fund at 17, and spent nights reading 10-Ks and pricing options by hand. Later, I built AI-as-a-service for financial firms for eight years. That work taught me two truths: institutions run on signals buried in messy data, and very little of that insight ever makes it to everyday investors in a usable way. Prospero exists to close that gap.
- You started investing as a teenager, how did those early experiences shape the way you think about risk and opportunity today?
I came of age during the dot-com bubble, which meant my first lesson was humility. The principles stuck. Size positions so you can survive being wrong. Cut losers quickly, let winners prove themselves with data, not opinions. Focus on process over prediction. If the signals change, I change. No hero trades, no falling in love with tickers, just probabilities and risk math.
- What was the “aha” moment that inspired you to launch Prospero.ai?
There were two. First, our team beat a large global bank in a head-to-head modeling challenge using alternative data like news and social sentiment. The edge was real, and it was explainable. Second, I kept seeing how options flows foreshadowed big equity moves. That is when it clicked. We could translate institutional behavior into a handful of clear, stable signals anyone could use in minutes. Millions of data points, distilled into a simple scorecard, and a playbook people could actually learn.
- Prospero’s mission is to level the playing field between Wall Street and retail investors. What’s the biggest edge institutions still have that’s hardest to democratize?
Execution speed and access to bespoke order flow are the toughest. Hedge funds can pay for sub-second execution advantages or datasets that retail investors never see. But Prospero focuses on the most impactful institutional signals like options flows, dark pool trades, analyst revisions, and sentiment shifts, and distills them into a framework that is just as actionable for individuals. Scale will always favor institutions, but clarity and strategy can absolutely be shared.
- I’ve heard it mentioned that Prospero.ai has “100 million data points” and “10,000 AI models”. Can you walk us through what makes Prospero’s approach different from other retail trading apps?
We built Prospero to process the same data institutions rely on: options, dark pools, analyst revisions, news, and social sentiment. Every day we run more than 100 million data points through 10,000 proprietary AI models. Instead of overwhelming users, we translate all of that into ten simple signals, each scored from 0 to 100. That gives investors a clear picture of a stock at a glance. The result speaks for itself. Our newsletter paper trades have consistently outperformed the S&P 500, by 70 percent year-to-date in 2025, 77 percent in 2024, and about 50 percent in 2022 and 2023. Of course, signals are illustrative, past performance does not guarantee future results, and paper trading has its limitations. There are GenAI solutions that are even more simple, but when you ask it to compare stocks they will use old data like P/E ratios. And there are other effective tools out there but they often require desktop use and have long learning curves. Prospero is unrivaled in the union of simplicity and efficacy.
- What do you see as the long-term vision for Prospero.ai? Will it always be an investor tool, or could it expand into other areas of financial decision-making?
Prospero is designed as a financial copilot. Today that means stock and ETF signals, education, and soon, trade alerts. The roadmap extends into managed strategies, ETFs, model portfolios, and eventually tools that connect market signals to real-life decisions like retirement glide paths or cash management. The promise is consistent: take complex inputs, deliver simple actions, and teach as we go.
- Looking back on your career, what’s the most unconventional bet you made that turned out right? And what’s one that didn’t?
Starting with the bad bet—because it eventually led to the good one. At my last company, we bet big on complex non-linear equations. We were early to advanced technology, developed jointly with NYU, such as evolving neural networks (genetic algorithms paired with deep networks).
The system tested well at a high level, but we had no real understanding of how it was “thinking.” That lack of transparency caught up with us when we flopped in front of a major international asset manager during an equity-picking demo. What we didn’t realize at the time was that the system was actually predicting volatility, not price. We only discovered this after seeing several high-risk picks perform poorly in the POC. Four of the ten picks were in biotech — and shortly after we presented, the sector had a rough couple of months.
That was nearly a decade ago, but I still see many people in AI today overlooking the same issue. My concern is that it will end up costing a lot of people money in the next downturn.
The bet I got right came from learning that earlier lesson. First, I knew I needed to understand exactly why the models were making their choices. At the same time, I realized that in my pursuit of the most advanced non-linear equations, I was never going to catch up to firms like Bridgewater or Renaissance.
That realization led to the foundational models behind Prospero. I went back to “school,” starting from intro-level Python all the way through deep learning. But it wasn’t enough just to understand how the models were thinking about the problem—I needed to go further and understand how their “brains” actually worked.
So I made a big bet: if I could build foundational models, and simplify the complex non-linear into linear, I could create a system that was both faster and more intuitive. This approach allowed me to invent my own reinforcement learning process, where I could verify ground-level concepts like “Net Options Sentiment.” For example, does my 0–100 scale for this output reflect how I would personally rank stocks in the options markets when I review their chains?
Once we had strong, unique foundational models that explained exactly what I wanted them to, we could then layer on what I’d call “intelligent scale.” That gave us the best of both worlds: logic systems grounded enough that I could always verify the market comprehension, and AI-driven flexibility that dynamically adapts to changing conditions—for example, by adjusting the weights of our many foundational models.
- You’ve spent over a decade working with AI in finance. What excites you most about the current wave of AI breakthroughs?
Two things. First, the rise of smaller, faster, more interpretable models. That is a perfect fit for finance, where transparency matters. Second, real-time agents that can monitor, explain, and act within guardrails. Pair those with rigorous data pipelines, and you get personalized guidance that is both powerful and teachable. The future is not black box calls, it is transparent copilots that show their work.
- How do you typically hire talented team members?
We look for teachers and builders. Every candidate completes a practical exercise that mirrors our day-to-day, ships a small artifact, and writes a one-page explainer a non-expert could follow. We want curiosity, clean thinking, and respect for users. If you can simplify complexity without losing truth, you will thrive here. Beyond that, a lot of our early team worked for equity only. We are a mission-first company and that has ensured we have dedicated people with aligned goals.
- What’s the hardest leadership challenge you’ve faced scaling Prospero.ai, and how did you overcome it?
Focus. Along the way, we had opportunities to chase features, partnerships, and shiny objects. We made the hard call to cut nice-to-have initiatives, double down on signal quality, education, and alerts, and reset our operating cadence around weekly user outcomes. Clear priorities, honest postmortems, and shipping on a drumbeat turned the corner.
- If you were 17 again, just starting out in finance, how would you invest your first $10,000 today?
Stock picking is the future in my estimation and we positioned in my opinion, the best product for that. But for a beginner, I would lean on the risk control aspect of ETFs mixed with what I think are the surest bets, the cloud providers. I say risk-controlled because I like ETFs like QTUM and BLOK because you don’t have to pick a winner in Quantum or Blockchain, which is hard for experts, let alone novices. But also some safer things so this is how I’d break it down. 10% AMZN, GOOG, and MSFT, 10% BLOK, 10% QTUM. 25% SCHD to enhance the earning power, 25% VTI diversification.
- When you’re 80 looking back, what do you want to have accomplished in your life?
I want to be able to say we moved the needle on financial literacy at scale. That millions of people felt calmer and more capable because Prospero turned noise into clarity. I want to see a global community, strong education programs, and products that started with signals and grew into portfolios, funds, and planning tools people trust. We are already expanding internationally with new advisors, accelerator support, and partner programs in motion. Most of all, I hope alumni and users can say this work made their lives better, not just their returns. Beyond that, I hope to touch the lives of a great many people one on one and for people to say that I was someone who really cared about them. Many of our users already say this but I hope that I never lose that part as we grow.
We hope you enjoyed this interview. If you would like to be featured in a future interview, please contact our team at innovationinsider@donestreet.com.
To learn more about DoneStreet and why our customers love us, follow this link.
Spend Less Time Hiring, More Time Building. Hire With Confidence Through DoneStreet.
If you have not yet subscribed, please hit the subscribe button and let us know what types of content you’d like to see more of in the future.
Thanks for reading Innovation Insider - Presented by DoneStreet! Subscribe for free to receive new posts and updates.
