Q1. Could you start by giving us a brief overview of your professional journey, particularly focusing on how your experience in cloud architecture, agile delivery, and AI adoption has shaped your approach to enterprise technology transformation?

I began my career back in 2000, which means I’ve been in this field for nearly 25 years now. My journey has been what you might call “organic growth”—I began as a programmer and gradually moved up the ladder: senior developer, tech lead, delivery manager, senior delivery manager, and eventually head of department. But even as I took on more leadership roles, I always kept one foot in technology. No matter my title, I made sure to stay at least 50% hands-on. That technical curiosity has always been a driving force for me.

After COVID, I took on the role of a full-scale technical architect, handling not just the technical side but also the commercial aspects across different domains. My core experience has been in travel, hospitality, and especially aviation—I’ve spent more than 15 years focusing on aviation technology. Over the past decade, I’ve led several cloud transformation and application modernization projects, which really opened my eyes to both the possibilities and the challenges of large-scale change.

In the last two and a half years, I’ve seen AI become a much bigger part of what we do. We’re moving beyond traditional software development, actively integrating AI into our products. So, modernization and automation have become central themes. Currently, I work as a Principal Solution Architect at a major airline, where I’m responsible for three big programs, including one focused on generative AI. I lead a team of 15-plus architects. That mix of hands-on technology and leadership has really shaped how I look at enterprise transformation—it’s about balancing innovation with practical delivery.

Q2. With enterprises rapidly adopting generative AI, what architectural patterns or data strategies do you see as most effective in ensuring AI solutions deliver real business value rather than hype?

The AI landscape has changed dramatically since generative AI solutions like ChatGPT came on the scene. These days, organizations are much more focused on cost efficiency and outcome-based development. The focus has shifted from building features for the sake of it to solving real problems and delivering measurable value. Now, before we even start building, the first questions are always: “What’s the ROI?” and “What’s the actual business impact?”

Key strategies include:

• Analysing type if enterprise data and volume data (Input/Out) 
• Prioritizing cost efficiency and real business outcomes
• Embedding security, compliance, and regulatory requirements from the start
• Bridging the customer trust gap using customer data through using AI, such Gen or Agentic AI solutions, education and involvement 

What I’m seeing is that many clients approach AI cautiously at first. Generally, their trust level in generative AI is around 25–35% when we start. It’s our job to bridge that gap, educating and involving them until they’re comfortable moving forward, usually around that 60–70% trust mark. That education piece is critical for turning AI from hype into something that genuinely moves the needle for the business.

Q3. Given your multi-cloud expertise, how should large organizations balance cost, resilience, and innovation when deciding between single-cloud versus multi-cloud strategies?

That’s a great question. Even today, a significant chunk—around 35%—of enterprise customers still keep their data on-premises. Often, it comes down to security concerns or just wanting to keep their data close to home. In those situations, we’ll use models and tools that are optimized for on-prem, like Hugging Face with custom tokenization to keep the compute costs manageable.

But cloud has completely changed the game. Platforms like AWS, Azure, IBM, and GCP have made it possible to design, develop, and deploy AI solutions at scale without worrying about the infrastructure. AWS Bedrock and SageMaker, for example, let you spin up AI models quickly and cost-effectively. That said, cloud isn’t the answer to everything, but it certainly removes a lot of traditional blockers.

A big trend I see is that large enterprises don’t want to be locked into a single cloud provider. So, they’ll often use two or even three cloud providers—like AWS and Azure together, or Azure with GCP. This multi-cloud approach helps avoid vendor lock-in and lets organizations pick the best tools for their needs.

Personally, I take a cloud-agnostic approach. My role as an architect is to be flexible—use whatever cloud or combination of clouds fits the client’s goals, budget, and data requirements. That way, you strike a balance between cost, resilience, and innovation.

Q4. How is the rise of AI-driven cyberattacks changing the way architects approach DevSecOps, identity management, and “security by design” in cloud-native systems?

To be honest, AI-driven cyberattacks have become part of daily life for any organization with public-facing apps. Most companies deal with at least 10–12 medium-level attacks every single day, and usually one or two serious ones. Generative AI has definitely made things trickier—prompt injection and jailbreaking are much easier now, and even with strong guardrails, persistent attackers can sometimes find a way through.

Because of this, security can’t be an afterthought anymore. “Security by design” means thinking about things like role-based access, least privilege, and strong guardrails from the start. In the cloud, AWS and Azure provide real-time dashboards for monitoring model health and security, and these tools have become indispensable for us.

Continuous monitoring is key. We keep a close eye on prompt logs and user behaviour to spot anomalies early. While you can’t eliminate all risks, these proactive steps help keep vulnerabilities to a minimum and ensure that any issues are caught and dealt with fast.

Q5. You’ve emphasized architectural governance and the use of ADRs. How do you ensure agility and innovation while still maintaining consistency and accountability across global teams?

Governance in large organizations—especially those with 50,000–60,000 employees—can be a real challenge. When new people join, they want to bring their own tools and ways of working, which is great for innovation but also increases the risk of inconsistency or technical debt if not managed well.

That’s where Architectural Decision Records (ADRs) come in. We use ADRs at both the project and technology levels. Project-level ADRs are reviewed by senior or principal architects and cover incremental design changes. Technology-level ADRs are for the big decisions, like switching from Oracle RDS to DynamoDB or adding Google Cloud to our stack for multi-cloud support. These require sign-off from the heads of architecture.

Once a technology-level ADR is approved, it gets shared across the organization so other teams can leverage the research and decisions that have already been made. We also maintain a tech catalogue listing all approved software, libraries, and AI services (for example, AWS Q for Java migration projects). This approach saves time and ensures everyone is on the same page.

So, by formalizing decision-making through ADRs and sharing that knowledge, we keep things agile and innovative while maintaining much-needed consistency and accountability.

Q6. Airlines and travel companies are experimenting with digital wallets, blockchain-based loyalty, and AI copilots. Which of these technologies do you see becoming mainstream in the next few years, and why?

Right now, digital wallets are leading the pack in the travel and airline industries. I see them going mainstream in the next four to five years. Blockchain had its moment a couple of years ago, especially for things like secure parameter storage and secret management, and while it’s still valuable for highly secure scenarios, it’s not the primary focus right now. AI copilots are starting to gain traction, but their mainstream adoption will depend on how the regulatory and compliance landscape evolves.

In practice, many organizations are integrating digital wallet infrastructure with cloud-based tools like AWS Secret Manager and Parameter Store for secure credential management. While I haven’t personally driven a digital wallet implementation, at the program level we rely heavily on AWS services to securely store secrets, credentials, and connection details.

To sum it up: digital wallets are set to become widespread first, with blockchain and AI copilots not far behind as security and automation needs grow.

Q7. If you were an investor evaluating companies undergoing cloud and AI-driven digital transformation, what critical question would you ask their leadership about future readiness and competitive differentiation?

The first thing I’d want to know is: What’s the ROI? Whether it’s a new product or a major transformation, leaders should be able to articulate the tangible value—be it revenue, cost savings, or efficiency gains—that they expect to see, even at the MVP stage.

Next, I’d ask: How does this initiative reduce manual work and streamline legacy processes? The goal isn’t just to add more technology for the sake of it, but to actually simplify and improve operations.

Finally, I’d probe into how these changes will impact overall system reliability and security. In large enterprises, with hundreds of applications running simultaneously, it’s crucial to understand how new solutions will stabilize the ecosystem and mitigate risks.

So, the critical checks are ROI, operational efficiency, reliability, and security. If leadership can answer those clearly, it’s a good sign they’re ready for the future and have a real edge over the competition.