On September 25, 2025, Nanyang Technological University in Singapore hosted a highly engaging webinar titled “Coding the Markets: AI Innovations and FinTech Career Insights.” The event, attended by NTU graduate students, featured three renowned speakers – Jiri Pik, Jared Broad, and Ernest (“Ernie”) Chan – who shared valuable perspectives on technology and careers in quantitative finance. (All three speakers are co-authors of the Wiley book Hands-On AI Trading with Python, QuantConnect, and AWS, a comprehensive guide with 20+ practical AI trading strategies) This recap distills each speaker’s key messages and the core takeaways for aspiring fintech professionals. (Slides from Jiri Pik’s talk are available for download – see Coding the Markets Slides here.)
Jiri Pik: Thriving in an AI-Driven FinTech Career
Jiri Pik, a veteran quant and author, opened with a reality check on breaking into finance. Entry-level roles in trading and quant research are extremely competitive – investment banks’ graduate program acceptance rates are approaching 0.1%. In other words, “impossible to get in” by conventional means unless you attend a target school, have inside connections, or otherwise stand out by solving a problem the hiring manager cares about. Jiri emphasized that GenAI (Generative AI) is further shrinking junior job opportunities, automating many tasks that newcomers used to do. Citing a recent Stanford study, he noted that AI-driven automation has already cut about 6% of jobs in AI-exposed fields like coding, disproportionately affecting young professionals. Yet there’s a flip side: those same advances create demand for a new kind of talent. “AI-native” graduates – those skilled in AI tools and workflows – are highly sought after; in fact, some 20-somethings fluent in AI are commanding six-figure salaries right out of school. In Jiri’s words, “The AI-skilled grads are welcome,” even as traditional roles disappear.
Despite these challenges, Jiri’s message was inspiring: you can break in and thrive by out-hustling and differentiating yourself. He urged students to focus on measurable accomplishments and practical experience rather than generic credentials. In an ultra-competitive landscape, having a “non-standard résumé” is key – one that showcases real projects, code, and results that matter to employers. Publish a case study, contribute to open-source, share a project on GitHub or a blog; this not only proves your skills but also builds your personal brand in the industry. Networking remains critical (“be heavily networked” as Jiri put it). Many jobs aren’t even advertised publicly, so connecting with insiders and mentors can open doors that online applications won’t. Jiri echoed a piece of wisdom making rounds on quant forums: you must “learn to suffer successfully.” In other words, be resilient and persistent. Rejection and adversity are inevitable in this field – what separates those who eventually succeed is the ability to endure, learn, and keep improving. As one hedge fund quant leader famously said, “You don’t learn how to succeed… you just learn how to suffer successfully.” Jiri highlighted this stoic mindset to remind young quants that perseverance and continuous learning matter enormously.
Importantly, Jiri also stressed balance between cutting-edge innovation and robust execution. To stand out, you do need to be at the frontier of what’s technically possible – “be AI-native,” as he quips. Embrace the latest tools: generative AI coding assistants, cloud computing, and automation. These can make you a far more productive researcher or developer (2–5× faster by some estimates) and enable even a small team or individual to “ship institutional-grade infrastructure” quickly. Jiri advised mastering modern AI workflows (prompt engineering, retrieval augmented generation, model evaluation, etc.) and achieving cloud fluency – from batch processing to serverless architectures – so you can leverage unlimited compute and data on demand. Being adept with platforms like Azure and using AI assistants for tasks like code generation and testing will greatly amplify your output.
At the same time, technical prowess alone is not enough. Jiri cautioned that with great power (of advanced tech) comes great responsibility: you must rigorously manage the risks of bugs, data errors, and fragile infrastructure. In live trading, a single glitch can be costly. The hardest problems in quant finance often boil down to judgment and reliability, not just algorithm design. Knowing which model or technology to apply, and ensuring your code and data pipelines run flawlessly day in and day out, is more valuable than inventing the fanciest new model. Jiri pointed out that truly world-class quants set themselves apart by delivering high-quality, low-defect systems – they “raise the bar” on reliability and define best practices. In other words, anyone can experiment with AI, but the winners are those who can consistently execute without blowing up. This theme of combining innovation with caution would resurface throughout the webinar.
To summarize Jiri’s career advice, here are some key strategies for aspiring quants that he and other industry leaders recommend:
- Differentiate Yourself: Traditional entry paths are drying up, so find alternate ways in. The hiring market is ultra-competitive – acceptance rates for finance grad roles are near zero. Stand out by showcasing unique achievements or solving real problems that matter to employers, rather than just degrees or test scores.
- Build a Non-Standard Résumé: Go beyond the basic CV. Create a portfolio of shipped work – e.g. coding projects, trading algorithms, research publications, or hackathon wins. Hiring managers love to see tangible proof of skills. If your résumé highlights “achievements of interest” (with metrics and results), you’ll command attention.
- Network and Use Referrals: Many quant jobs are filled through headhunters or personal networks, not public job postings. Leverage LinkedIn, alumni, professors, and events to form connections. Often, who you know can get you an interview more reliably than blind applications. Even top firms hire via recruiters – one quant lead got his break at Citadel through an external referral. Don’t be shy about reaching out.
- “Learn to Suffer Successfully”: In this field, resilience is essential. You will face steep learning curves, intense competition, and occasional failure. Take each setback as a lesson rather than a verdict on your ability. As Balyasny’s Head of Quant Research advised interns, “Finance is built on being right just a bit more than being wrong… You just learn how to suffer successfully.” Embrace challenges and stay in the game – perseverance pays off.
- Embrace Niche Expertise: Consider developing skills in a niche domain where you can excel. This could mean specializing in an emerging market, a specific asset class, or a cutting-edge technique. For example, given Asia’s growing financial markets, Jiri suggested gaining deep knowledge of Asian market microstructure, positioning yourself to ride the wave of new liquidity and innovation. Being an expert in a less crowded area can make you invaluable.
- Be “AI-Native” and Cloud-Fluent: Modern finance is increasingly tech-driven. Master the AI and cloud tools that can accelerate your work. From using GPT-style coding assistants for rapid prototyping, to harnessing cloud platforms for scalable backtesting, technology can be your force-multiplier. The most successful new quants are those comfortable with AI APIs, machine learning libraries, and cloud services. (In fact, Jiri noted that platforms like QuantConnect are a perfect training ground – it “remains one of the easiest ways to learn by doing” in algorithmic trading.) Continuously adopt tools that boost your productivity and keep you at the cutting edge.
Ernest P. Chan: From Predictive to Corrective AI in Trading
Dr. Ernest (“Ernie”) Chan, a pioneer in algorithmic trading and founder of PredictNow.ai, brought a fresh perspective on how AI is applied in trading strategies. Ernie’s core message: AI that corrects human decisions is often more effective than AI that tries to predict the market outright. He began by drawing an analogy: fully autonomous systems have often underwhelmed (think of self-driving cars’ “unfilled promises”), whereas assisted intelligence is everywhere (driver-assist features are now standard). The same, Ernie argues, holds in finance. Purely predictive trading algorithms – ones that attempt to pick every stock or time every market move – struggled for many years. In fact, Ernie reminded us of the “Financial AI Winter” from roughly 2000–2018, when many machine learning trading funds failed to live up to the hype. The reasons were manifold: overfitting to past data, reflexivity (markets adapting to the models), and sudden regime changes that broke models’ assumptions. Simply put, predicting prices is hard.
However, Ernie found that using AI in a corrective, risk-management role yields far more consistent success. Instead of predicting the market’s every twist, his approach (implemented in PredictNow.ai’s cloud platform) focuses on predicting and managing risk. He calls this “Predictive Risk Management” or Corrective AI. The idea is to continuously learn from your own strategy’s performance – identify when your strategy is likely to hit a losing streak or when its risk is spiking – and then automatically adjust parameters to mitigate losses. In practice, this means employing machine learning to optimize portfolio allocations or strategy parameters (a process Ernie terms Conditional Portfolio Optimization, CPO) using your strategy’s historical data and a rich set of features. At the same time, the AI analyzes which factors (features) are driving your strategy’s risk – providing explainable insights into why the strategy might fail under certain conditions. Finally, the system computes a forward-looking probability of loss for the next period for your strategy – essentially a dynamic risk forecast. If the predicted risk is too high, the “corrective AI” can signal you to reduce positions or hedge. In Ernie’s words, this is about using ML not to tell you “what to buy,” but to tell you “when to pull back.”
The results Ernie shared were compelling. He presented case studies where this corrective approach significantly improved performance metrics. For example, in one equities portfolio, the CPO algorithm automatically shifted to defensive, cash-heavy positions ahead of a bear market, thereby reducing drawdowns and improving the Sharpe ratio by over 50% compared to a static allocation. In another strategy example, CPO optimization boosted the Sharpe ratio by more than 100% versus a traditional equal-weight portfolio, with substantially lower volatility and smaller drawdowns to boot. These are remarkable gains achieved not by forecasting higher returns, but by avoiding losses and rebalancing wisely. Ernie humorously noted that many quants spend their days trying to predict the next market jump, when sometimes the bigger win is simply not jumping off a cliff. His “Corrective AI” framework essentially watches for cliffs and guides you away from them.
What does this mean for those pursuing FinTech and machine learning roles? Ernie’s talk underscored that real-world success demands a blend of solid statistical know-how, strong programming, deep domain knowledge, and plenty of hands-on experimentation. There is no single algorithm that guarantees riches; rather, it’s the integration of good techniques with sound judgment that yields reliable profits. He advised students to ensure they have a firm grasp of statistics and probability (to avoid fooling oneself with noise), to become proficient in coding (so they can implement ideas efficiently and correctly), and to learn the domain context of finance (so they understand the economic meaning behind model signals). Only with all three – math, coding, domain expertise – plus the experience earned by actually trading or simulating strategies, can one truly appreciate what works and what doesn’t. As Ernie demonstrated, an AI model’s output is only as useful as the human wisdom that interprets and applies it. In the end, he echoed a unifying insight of the webinar: judgment is the secret sauce. Knowing when to trust the model, when to override it, and how to keep your trading system robust under stress is what separates top-tier quants from the rest.
Jared Broad: Democratizing Algorithmic Trading at Scale
Jared Broad, CEO and founder of QuantConnect, took the audience on a journey from his personal story to the global impact of cloud quant platforms today. Jared recounted how, back in 2008, he was a 20-something starting a small hedge fund out of his apartment. Those early years were a grind – he spent countless nights gathering data, coding infrastructure from scratch, and searching for trading signals, sacrificing any semblance of a social life. After a couple of years managing friends-and-family money, he experienced firsthand the “enormous barriers to entry” in quantitative finance. In 2011, that frustration catalyzed the birth of QuantConnect. His vision was to open up algorithmic trading by providing talented individuals the tools and data access typically available only at top hedge funds.
Fast forward to 2025, and QuantConnect has grown into a thriving global ecosystem. Jared shared some eye-opening numbers: the platform’s community now exceeds 400,000 members worldwide, ranging from students and independent quants to professionals at major funds. Over 500 hedge funds and investment firms actively use QuantConnect’s technology, representing client assets under management from half a million to $35 billion. In aggregate, users have created hundreds of thousands of trading algorithms on the platform, and live strategies running through QuantConnect execute about $45 billion USD in trades every month. This massive scale – tens of thousands of cloud-hosted algorithms and connections to dozens of brokerages – shows how far the quant community has come. What once took Jared years of solitary effort in an apartment can now be done in minutes on QuantConnect’s cloud: “Millions of backtests” can be run cheaply using elastic compute, and live strategies can route to 25+ broker integrations with data feeds spanning U.S. equities, futures (CME), FX, and even crypto. QuantConnect has essentially democratized algorithmic trading, enabling anyone with an idea and coding skills to test and deploy strategies at scale.
Beyond the impressive growth, Jared emphasized how this platform lowers the barrier for newcomers to gain real, reproducible experience. Traditionally, to learn quantitative trading, one needed to cobble together data sources, set up servers, and be inside an institution to see real market plumbing. Now, a student can log in to QuantConnect and immediately access institutional-grade tools: a clean IDE, extensive datasets, cloud computing, and a community for support. This allows for an education by doing – you can develop strategies, backtest them on decades of data, and even run them live with minimal setup. Jared pointed out that this kind of hands-on practice is invaluable for career development. Every project on QuantConnect is a potential résumé highlight or research paper. Moreover, by using a common platform, results are reproducible and scalable – an algorithm that works in backtesting can be taken straight to live trading on the same system, which closely simulates real-world conditions. As Jiri also noted in his blog, QuantConnect provides an “ideal playground” for mastering algo trading: it has an intuitive interface, ready compute resources, and free access to key market data – an unparalleled environment for both beginners and pros. In short, what Jared has built with QuantConnect is not just a business, but a training ground empowering the next generation of quants.
Finally, Jared plugged the new book “Hands-On AI Trading with Python, QuantConnect, and AWS” – a collaborative effort between himself, Jiri Pik, Ernie Chan and others – which encapsulates much of this knowledge. He highlighted that the book contains 19 fully implemented AI trading strategies that readers can literally copy, run, and modify on the QuantConnect platform. These strategies span hot topics like natural language processing for news sentiment, deep learning models for price prediction, and reinforcement learning for hedging. By studying these real examples (with complete Python code provided), readers can learn cutting-edge techniques and see how they perform in various market conditions. Jared’s point was that this resource gives aspiring quants a huge head-start: instead of reinventing the wheel, you can build on proven templates and focus on understanding why and when certain AI approaches work. (For more details on the book’s content and origin, see Jiri’s earlier blog posts introducing the project and the blog post announcing its release. And if you’re interested, you can find the book on Amazon.) Overall, Jared Broad’s talk left the audience inspired about the possibilities of cloud-powered finance. QuantConnect’s story – from a solo passion project to a platform supporting an entire global community – exemplifies how innovation can level the playing field. It also reinforced a practical takeaway: for anyone wanting a FinTech/quant career, there’s no excuse not to start building strategies today. Tools and knowledge that were once closed off are now at your fingertips.
Beyond Models: The Power of Judgment and Reliability
A unifying theme emerged from all three speakers’ insights: in the real world of quantitative finance, the hardest problem isn’t building the model – it’s knowing which model (or technology) to build and making sure it runs reliably. As exciting as AI innovations are, success ultimately comes down to human judgment and operational excellence. Jiri Pik drove this point home when he cautioned young quants not to chase every shiny object without a plan; it’s critical to discern which techniques actually add value versus which are academic exercises. Jared Broad’s journey illustrated that having the latest AI strategy means little if you lack the infrastructure and discipline to execute it consistently. And Ernie Chan showed that a modest strategy with rock-solid risk management can outperform a brilliant strategy that’s fragile. In summary, models and algorithms are tools – powerful ones – but it’s the craftsman who determines the outcome.
The panelists encouraged budding professionals to cultivate a sort of technological wisdom: combine curiosity about new advances with healthy skepticism and a focus on robustness. For instance, don’t just develop a complex machine learning model and assume it will print money. Test it in different regimes, check its failure cases, and think about how it might break if market conditions change or if data feeds glitch. Finance is rife with stories of the “perfect” model that worked until it didn’t, causing massive losses. The antidote, as our speakers stressed, is a relentless commitment to quality control and reliability. Even in the age of cloud computing and autoML, there is no substitute for carefully checking for bugs, ensuring data integrity, and designing systems with backups. In fact, one of the key insights from this webinar was that operational reliability often matters more than raw predictive power. An algorithm that’s right 55% of the time but never fails catastrophically will likely beat one that’s right 60% of the time but occasionally blows up due to an overlooked issue.
In practice, this means aspiring quants should invest as much effort into engineering and validation as they do into modeling. Adopt best practices: code reviews, version control, rigorous backtest validation (to avoid overfitting), and ongoing monitoring of live strategies. This might not sound as glamorous as designing a fancy AI, but it’s what professionals do to earn trust and survive in the market long term. As Jiri noted in his slides, top-tier quants differentiate themselves by shipping high-quality, reliable work that sets industry benchmarks. That reliability builds credibility – whether you’re pitching a strategy to a hedge fund or deploying an app to thousands of users. It’s ultimately a form of respect: respect for the complexity of markets and for the fact that real money (often other people’s money) is on the line.
In conclusion, the “Coding the Markets” webinar delivered both a dose of reality and a jolt of inspiration. The reality is that breaking into quant finance is challenging and the field is evolving rapidly with AI – complacency or a narrow skill set will not suffice. But the inspiration comes from knowing that the opportunities have never been greater for those willing to adapt and push themselves. As the speakers showed, an enterprising grad student today can leverage open platforms like QuantConnect, cloud AI services, and community knowledge to do things that a generation ago required a Wall Street firm. The keys are to stay curious, keep learning, and engage actively – whether by networking with mentors, collaborating on projects, or even co-authoring research. And when you do land that role or launch that fund, remember the deeper lesson: blend innovation with judgment. The quants who thrive will be those who not only code the markets, but also understand the markets – and never stop refining both their models and themselves.
Download Slides: If you’d like to explore more, click here to download Jiri Pik’s full slide deck from the event, which contains additional insights and references.
