Q1. Could you start by giving us a brief overview of your professional background, particularly focusing on your expertise in the industry?

I’ve spent over 26 years in software engineering, and over the last decade, my focus has been on leading large, globally distributed delivery programs. My experience sits at the intersection of engineering execution and business outcomes, particularly in customer operations and enterprise platforms.
I’ve led multi-location teams across regions, managing end-to-end delivery, vendor ecosystems, and senior stakeholder engagement. Over time, my role has evolved beyond just delivery into ensuring adoption, stability, and measurable business value.

More recently, I’ve been working extensively in the applied AI space, leading early implementations of agent-based systems and multiple proofs of concept across emerging platforms. The focus has been on taking AI from experimentation into real, production-ready solutions that can scale.

 

Q2. With data center energy demand hitting record highs in 2026, how much of the technical roadmap is currently gated by 'grid access'? Specifically, what is the financial risk to the margins if we see a significant spike in electricity costs?

Grid access is no longer just a background dependency; it’s becoming a key factor in shaping technical roadmaps, especially with the rise of AI-driven workloads. In some regions, infrastructure expansion is already being constrained by power availability.

From a financial standpoint, energy cost volatility directly impacts margins. For compute-intensive workloads, even a moderate increase in electricity pricing can significantly raise operating costs.

 

Q3. Given your delivery across the UK, US, and India, how are 'Data Residency' laws in 2026 impacting the ability to use a single global AI model? Does the need for 'Localized AI' significantly increase Total Cost of Ownership (TCO)?

Data residency is fundamentally changing how global AI systems are designed. The idea of a single, centralized model serving all regions is becoming less viable for large-scale deployments.

We’re seeing a clear move toward federated or region-specific models, either trained locally or fine-tuned within jurisdictional boundaries. That naturally increases TCO, not just due to duplicated infrastructure but also due to added governance, compliance, and lifecycle management overhead.

In countries like the US and Germany, where data privacy regulations are stricter, the challenge goes beyond architecture. The time required for DPIA approvals and legal alignment can significantly slow deployments. What stands out is that while AI agents can be built quite rapidly, getting them into production becomes the real bottleneck due to these compliance processes.

So, the challenge isn’t the speed of innovation, it’s managing regulatory complexity without slowing down delivery momentum.

Q4. In the context of global customer operations, are you seeing a shift toward 'Edge Intelligence' to bypass central cloud costs? How does moving processing to the 'edge' change the long-term infrastructure CapEx and hardware replacement cycles?

Yes, especially in customer operations where latency and cost efficiency are critical, there is a growing shift toward edge intelligence. Organizations are moving certain processing closer to the source instead of relying entirely on centralized cloud systems.

This helps reduce recurring cloud costs and improves responsiveness, but it also shifts the financial model from OpEx-heavy to higher upfront CapEx in edge-capable infrastructure.

At the same time, with AI workloads, token consumption is emerging as a significant cost driver. Both input and output tokens need to be closely monitored, as costs can scale quickly if left unchecked. There’s also a risk angle: poorly governed agents can be exploited to drive excessive token usage, which can affect both costs and system stability.

Moving to the edge doesn’t remove these concerns, but it changes how they are managed. Strong governance around workload execution and cost monitoring becomes essential.

 

Q5. In this volatile 2026 market, what is the specific signal that tells you to stop 'Building' native software and start 'Buying' SaaS? Has the rapid evolution of AI made custom-built software a financial liability rather than a proprietary asset?

It really depends on the use case and the maturity of the SaaS ecosystem in that area.

If you’re already invested in a strong platform ecosystem, and it offers capabilities to rapidly deploy solutions, especially AI-driven features like agents, it often makes sense to leverage that for speed and integration. But cost is something you need to keep a close eye on, as these platforms can quickly scale up expenses.

On the other hand, if you’re working within a more flexible engineering ecosystem, you may have the option to combine low-code and pro-code approaches, giving you more control and customization.

So, the decision is less about “build vs buy” and more about alignment on how well the platform fits your current ecosystem and where you want to go.
Another important shift is the planning horizon. The traditional 7–10-year technology strategy no longer holds. Given the pace of change, systems can become outdated in 2–3 years.

So, the focus now is on being agile, thinking in 2–5-year windows, continuously reassessing, and being ready to pivot. The organizations that do well are the ones that stay flexible rather than locking themselves into long-term, rigid decisions.

 

Q6. In 2026, 'Agentic AI' systems often have broad permissions across enterprise silos. What specific control protocols can be implemented for an autonomous agent that begins making unauthorized financial commitments or leaking PII through 'Prompt Injection'?

In practice, we’re seeing that building agentic AI for complex workflows is relatively straightforward, whereas deploying it safely within an enterprise environment is not.

The real challenge starts when these agents operate across systems on evolving platforms. There are no well-established standards yet for IT control frameworks or AI-specific governance, so a lot of this is being defined from scratch.

From a control’s standpoint, enforcing least-privilege access is critical, as is introducing approval layers for sensitive actions, such as financial commitments. Input and output guardrails are equally important to reduce the risk of prompt injection and unintended data exposure.

Auditability is another key requirement; every action taken by an agent should be traceable.

But beyond individual controls, governance is the bigger piece. Setting up monitoring, review mechanisms, and fallback strategies is essential. As these agents become more autonomous, the risk surface expands, and enterprise readiness becomes the main challenge and not the technology itself.

Q7. If you were an investor looking at companies within the space, what critical question would you pose to their senior management?

I’d focus on whether the organization has genuinely understood where AI creates real business value.

The question will be:

Have you clearly identified the use cases where AI improves productivity or accelerates time-to-market, and can you scale them effectively?

I’d also want to understand how the organization is responding internally. Are teams aligned with the long-term strategy? Is there real adoption, or is it still at a conceptual level?

And importantly, is the infrastructure ready? A lot of AI initiatives, especially agent-based ones,, are struggling not because of the idea, but due to gaps in non-functional readiness, such as scalability, reliability, and integration maturity.

So ultimately, it comes down to clarity of purpose, organizational alignment, and whether the foundation is strong enough to sustain AI at scale.