Georgios Koumantari: Real tech industry breakthrough won’t come from single technology but convergence of several

The real breakthrough in the tech industry won’t come from a single technology but from the convergence of several, according to Georgios Koumantari, HFM Group's Chief Technology Officer

As he says in the interview below, the true potential of artificial intelligence (AI) will unfold when it works in tandem with advances in quantum computing, blockchain and modern cloud infrastructure.

What are the specific AI tools that you have adopted? How well did they integrate with your existing infrastructure? Did they require specific customisation to work effectively?

We built an in-house anomaly detection system to enhance visibility across operational logs. Using AI models like logistic regression, support vector machines (SVM), naive Bayes, and random forests, we transformed log data with techniques such as TF-IDF and achieved high accuracy – SVM surpassed 97%, and random forest reached 92%. This system significantly reduced manual reviews, improved accuracy and saved time during audits and investigations across key workflows like payments and reconciliation.

Today, AI is deeply embedded in our operations. We use models from OpenAI and Google Gemini not just for productivity but within automated workflows. We see AI as a programmable layer that boosts efficiency and decision-making without compromising control. Beyond our custom solutions, we leverage built-in AI features in tools like Google Workspace and Atlassian Jira for tasks like document summarisation, translation and ticket resolution.

Have these AI deployments led to revenue growth, cost savings or improved customer retention? Can you quantify these improvements? And how long will it take before you see a measurable return on your investment?

Our AI strategy is focused on driving measurable business outcomes by deploying tools that deliver clear operational value – whether through cost savings, productivity gains or improvements in customer engagement and retention. Most initiatives demonstrate ROI within the first quarter and scale efficiently with minimal overhead. On the revenue side, our AI-powered transcription and conversation analysis platform, using OpenAI and Gemini models, analyzes client calls, extracts patterns, and generates actionable insights. This dual-provider setup ensures resilience and continuous performance. In fully deployed regions, we’ve seen a 10–12% conversion increase and an 8% drop in onboarding abandonment. Looking ahead, we expect up to 20% further gains in engagement, with added multilingual support, risk flagging, agent performance insights and predictive engagement scoring.

What have been the biggest bottlenecks in integrating AI solutions? How did you overcome these challenges?

The most persistent bottleneck when integrating AI at scale has been navigating technical and regulatory boundaries, especially around security and data protection. As a regulated financial organisation, we adhere strictly to privacy standards and ensure that every AI integration aligns with our internal security architecture. We operate on zero-trust principles, which means no system or user is assumed safe by default. Every AI interaction is audited, monitored and restricted within secure environments. This adds complexity to implementation but it’s non-negotiable when you’re dealing with sensitive client and transactional data.

As AI continues to transform the workplace, how has HFM adapted? How have employees responded and what steps have you taken to support this transition?

The shift is already underway, and it is accelerating. In my view, organisations that embrace AI will thrive. Those that hesitate risk falling behind. I see AI as a generational change in how work gets done. It is no longer optional. It is foundational to staying competitive, efficient and relevant. We’ve fully embraced this reality by creating new in-house roles, including AI software engineers and embedding AI into the fabric of how we operate. Rather than waiting for change to happen around us, we’ve made a conscious decision to lead it. We view AI not only as a tool but as a capability, one that sets the stage for how work will evolve in every department. Our employees have responded positively. What we’ve seen is genuine engagement, curiosity and a willingness to adapt. Teams across the company, from engineering and support to compliance and finance, are interacting with AI tools daily. They’re not just using them passively; they’re contributing feedback, sharing creative uses and shaping how these tools are implemented.

In your industry, what types of decisions can realistically be entrusted to an algorithm? And do you think that AI will ever be capable of handling the novelty and ambiguity of executive-level decision-making?

In the financial industry, and especially in the trading and brokerage space, AI has the potential to solve one of our biggest longstanding challenges: the human tendency to make decisions based on gut feelings or legacy thinking rather than data. The truth is, many decisions are still made in boardrooms based on instinct, hierarchy or inertia. That won’t be sustainable much longer. If implemented correctly, AI can help us shift away from this. It can backtrace years of financial history, analyse news cycles, trading activity, client behaviour and internal performance metrics and surface patterns that no human could ever detect alone. But the quality of the output depends entirely on the quality and structure of the data. This is where much of the industry still falls short. I strongly believe that companies that embed data observability into their systems, not just to detect failures, but to create actionable intelligence, will lead. In terms of decision-making authority, AI should not be seen as a threat. It doesn’t need to replace executive roles and it shouldn’t. But it should be treated as a trusted co-pilot that can process the data faster, with more context and consistency than any individual. Executives who embrace this mindset will outperform those who resist it. In the end, the question is not whether AI can make executive-level decisions but whether executives are willing to let go of outdated decision-making habits and trust the data they’ve already been sitting on.

Which AI trend or technology are you watching closely and which do you think are more marketing fluff than substance?

One area I am actively exploring is AI-powered observability agents. These are intelligent systems capable of monitoring infrastructure, databases, traffic flow and product-specific behaviour across a distributed ecosystem. Unlike traditional alerting tools, these agents can detect patterns, uncover anomalies, and identify systemic risks before they escalate. I see this as a powerful way to strengthen operational resilience and provide real-time traceability across complex environments. When it comes to trends that I approach with more caution, agentic AI systems stand out. The concept of autonomous agents handling multi-step or parallel tasks across complex environments is promising but I have yet to see a production-ready implementation that meets the reliability, scalability and operational rigor expected in enterprise settings.

I’m also sceptical of relying too heavily on Large Language Models for decisions that require consistency. These models are excellent at summarising,

creating and assisting with exploration, but they are not deterministic. The same question can yield different answers depending on context, timing or phrasing. That makes them useful in certain roles but not where repeatability or accountability is required.

What is the one AI breakthrough that you believe will be a game-changer for your industry?

The real breakthrough in our industry won’t come from a single technology but from the convergence of several. AI has already changed how we operate internally, helping us automate, optimize and personalise. But its true potential will unfold when it works in tandem with advances in quantum computing, blockchain and modern cloud infrastructure. We are also seeing the rise of the always-on financial world. The 24/7 nature of crypto has set a new standard. Whether or not you operate in digital assets, your clients now expect platforms to be available, stable and responsive around the clock. That pushes all of us toward cloud-native infrastructure. Systems built to be cloud-native from the start have a clear advantage. They support fast deployments, elastic scaling, fault tolerance and global reach without complex overhead.

• This article was part of GOLD magazine’s Cover Story in April. To view it click here