I'm an AI and Cybersecurity Leader with over 25 years of international experience spanning the United Kingdom, Singapore, and Austria. I've served as Chief Technical Officer, Senior AI Solution Architect, and Data Science Lead for organizations including BAE Systems Applied Intelligence, Raiffeisen Bank International, and Nevis Security, bringing a unique blend of technical depth and strategic vision to each role.
I architect solutions that balance security, scalability, and functionality. My expertise encompasses cloud-native architecture across AWS, GCP, and Azure, designing resilient microservices infrastructures, and transforming legacy systems into modern, high-performance platforms.
My architecture philosophy centers on three pillars: security-first design, functional excellence that delivers business value, and scalability that supports growth.
I combine AWS Solutions Architect certification with hands-on experience in data architecture, ETL pipelines, database optimization, and DevOps practices including CI/CD with GitLab.
I blend offensive threat detection, defensive architecture, and comprehensive security governance. I bring hands-on expertise in cyber analytics, fraud detection, network security, and implementing defense-in-depth strategies aligned with the MITRE ATT&CK Framework and NIST AI RMF.
I spearhead enterprise-grade AI strategy with focus on Large Language Models, Retrieval-Augmented Generation (RAG), and Agentic AI systems. My expertise spans the complete AI/ML lifecycle—from data preprocessing and model training to secure deployment and governance—using technologies including VertexAI, PyTorch, TensorFlow, Langchain, Google ADK, and Model Context Protocol (MCP).
My work demonstrates the intersection of AI/ML and cybersecurity through NLP techniques and anomaly detection in cyber threat analytics. I serve on the AI Governance board, ensuring compliance with the EU AI Act, and adopt best practices from the NIST AI RMF for risk management in high-stakes domains including finance, insurance, and healthcare.
With over 25 years of hands-on software development experience, I've built mission-critical systems for tier-one financial institutions, telecommunications providers, and enterprise clients worldwide. My development expertise spans Java, Python, SQL, and modern frameworks, combined with deep knowledge of software engineering best practices, code quality, and team mentorship.
My development approach emphasizes clean code, comprehensive testing, documentation, and knowledge transfer. I'm proficient in modern DevOps practices including Docker containerization, Kubernetes orchestration, GitLab CI/CD pipelines, and continuous integration/deployment workflows that enable teams to deliver high-quality software rapidly. My experience includes work at Siemens, THUS Plc./Cable and Wireless, Sungard, and Fujitsu Telecom, building telecommunications and enterprise software systems.
Clear explanations of key concepts and approaches
Solution architecture is the art of designing a reliable and robust system capable of supporting your requirements for the foreseeable future.
As a system vendor, a good solution architecture document forms the basis of your relationship and contract with the customer, clearly articulating the different implementation phases, delivery and payment milestones. It is the key reference document that engineering teams use as they go through the detailed implementation steps of the system.
As the customer, a well written solution architecture document clearly frames and describes your business requirements and use cases as well as your path to a good ROI. It is a clear map to your board of executives as to why the system being implemented is profitable and viable and should reduce red tape when it comes to regulatory review.
System architecture takes into account existing legacy systems and how they should be replaced or integrated with. It determines any regulatory or geographical factors to inform whether the system should be on-premises, on-cloud or a hybrid of both. It outlines the required scalability and elasticity, allowing you to use such modern technologies as Kubernetes to expand and contract your capacity on demand, allowing you to meet demand in busy periods as well as saving costs and resources in times of less demand.
Cybersecurity is the practice of defending all digital assets—systems, networks, and data—from digital attacks and unauthorized access. It deals with security threats across all vectors.
There are two main aspects to cyber security: "perimeter" and "post-perimeter" defense.
Perimeter defense concerns itself with keeping intruders outside of your internal network through the use of a firewall or other Intrusion Prevention and Detection Systems (IDPS). An IDPS will typically employ such methods as Rate Limiting (to stop denial of service attacks) and access control (denying known blacklisted IPs and allowing known IPs).
Post-perimeter defense concerns itself with what happens once an intruder has infiltrated an internal network. This will typically employ advanced cyber machine learning analytics to spot malware or suspicious unusual behavior and alert a Security Operations Centre (SOC) of a possible intrusion.
In today's modern cloud-based systems, a blended approach is required between perimeter and post-perimeter defense techniques. The tools and techniques to achieve this are available, but it is only through careful expert planning, implementation and regular reviews that you can achieve the high level of security needed.
Artificial intelligence (AI) is a field of computer science dedicated to creating systems that can learn, reason and solve complex problems in a way that is inspired by how a human might solve the same problem.
Machine Learning (ML) is a subset of AI that has been around for many years. ML models are used in many scenarios such as fraud and cyber attack detection and prevention. Common model implementations use regression techniques to learn patterns based on historic data such as what banking system activity could be indicative of fraud. These trained models are then fed real-time data in order that they can predict when an event such as fraud or a cyber attack can be detected.
Large Language Models (LLMs) are another form of AI that came to prominence in November 2022 with the arrival of OpenAI's ChatGPT 3.5. They are categorized by their ability to hold coherent conversations on a seemingly wide variety of subjects. There are several LLM vendors currently, most notably OpenAI (ChatGPT), Google (Gemini), and Anthropic (Claude).
LLMs give rise to many business applications such as chatbots—a bot that uses LLM technology and company documentation to talk about a company's product, for example. The other main business application is the coupling of LLMs with other tools to achieve Agentic AI. Agentic AI goes one step beyond the functionality of an LLM; it effectively extends the LLM's functionality so that it can interact with other systems and tools on your behalf to achieve a pre-defined goal. Although AI agents give rise to many business applications and savings, they come with risks if the proper security and guardrails are not put in place.
Contributing to the local tech and AI in diplomacy charter
Supporting initiatives to help children begin their journey into coding
Member of the TED-AI community advancing AI discourse
AI-assisted development for rapid prototyping and implementation
Professional design assets and visual elements
Modern web standards, responsive design, and best practices