Building Technology That Keeps Business Moving

Chirag Parikh

SVP of Technology – Head of Data, AI and Digital

Confiz

Building Technology That Keeps Business Moving

Chirag Parikh

SVP of Technology – Head of Data, AI and Digital

Confiz

Chirag Parikh-Most Innovative Tech Leaders from USA 2026

Innovation is rarely limited by the technology itself. More often, it is limited by how much freedom an organisation has to change. Chirag Parikh’s career has increasingly been shaped by that distinction, moving from complex financial systems and enterprise architecture into large-scale digital transformation, cloud, data and AI. His experience across Fannie Mae, Capital One, Walmart and GEICO exposed him to businesses where technology could never be treated as an isolated engineering exercise. Systems had to work at scale, withstand real-world complexity and keep evolving without putting the business at risk. That perspective now sits at the centre of Chirag’s work as Senior Vice President and Head of Technology, AI and Data at Confiz, where he is building enterprise AI platforms, multi-agent architectures and autonomous workflows designed to change how businesses actually operate. His thinking extends beyond the excitement around new models and tools, asking harder questions about data, legacy systems, governance, organisational readiness and measurable business outcomes. For Chirag, the most sophisticated architecture is not necessarily the most valuable one. The real measure is whether it gives an organisation greater freedom to move.

In this conversation, TradeFlock explores the experiences and ideas shaping Chirag’s approach to innovation, enterprise AI and technology leadership.

What experiences changed the way you think about technology leadership?

Looking back, some of the biggest lessons in my career came when technology stopped being primarily an engineering problem and became an enterprise constraint problem. Financial services taught me that innovation without trust has no value because technology decisions can affect money, risk and regulatory outcomes. Walmart then changed my understanding of scale. A solution is not truly scalable simply because it works in the cloud. It has to continue working across thousands of locations, millions of transactions and imperfect real-world conditions. GEICO added another dimension because we were modernising mission-critical platforms while continuing to serve customers every day. There was no pause button. Those experiences made me think less about whether technology is technically sophisticated and more about whether it gives the business room to move. The best architecture is not the most sophisticated architecture. It is the one that gives the business the greatest freedom to move.

What are organisations still missing when they try to move AI into the enterprise?

A lot of the hardest AI problems are sitting outside the AI stack. Enterprises can have powerful models, GPUs and agents, yet still struggle with fragmented data ownership, complex approval structures, duplicated processes and years of technology debt. Placing AI on top of those conditions can simply accelerate the wrong process. Work at Confiz has reinforced that lesson across retail, CPG and other industries, where the initiatives that moved toward production began with a business problem rather than a model. Inventory optimisation, media campaign management, sales intelligence and conversational analytics all required trusted data, workflow integration, governance and people who could act on the insight. AI becomes valuable when it moves from answering questions to changing how the business operates. A thousand AI use cases do not equal an AI strategy. The real measure is whether AI changes revenue, customer experience, risk, cost or the quality of decisions.

How should leaders make technology choices when both AI and legacy systems keep changing?

Technology leaders should stop looking for the winning AI model because there will not be one. Different workloads will require different models, which makes model optionality increasingly important. Accuracy, latency, privacy, security and economics all matter, but I would add another measure that businesses should pay more attention to: cost per successful business outcome. Cost per token may interest engineers, while cost per resolved claim, completed transaction or prevented fraud event matters to the business. Legacy requires the same thinking. Age alone does not make a system bad. You do not need to replace the legacy system first. You need to remove its ability to hold the business hostage. Progressive modernisation, clearer domain boundaries, APIs, events and separated data ownership can create business flexibility without turning transformation into a disruptive replacement programme.

Where do you see the next meaningful shift in enterprise AI?

The next shift becomes interesting when AI stops behaving like a chatbot and starts behaving like a digital workforce. A complex insurance claim, for instance, can involve document analysis, damage assessment, fraud evaluation, policy validation, payment decisions and customer communication. Coordinating those activities across systems and teams is where multi-agent AI can create a fundamentally different workflow. The opportunity extends well beyond insurance into supply chain, finance operations, software engineering and customer service. Yet greater capability should not automatically mean greater autonomy. Low-consequence, reversible actions can move quickly, while financial commitments, regulatory decisions and eligibility determinations require stronger controls. Autonomy must be proportional to consequence. Every enterprise agent needs an identity, permissions, policies, auditability and clearly defined boundaries, because intelligence alone should never determine how much authority a system receives.

What philosophy now guides you as technology keeps accelerating?

My leadership philosophy has become increasingly simple. Be uncompromising about purpose and flexible about everything else. Strategies will change, technology will change and AI will accelerate that cycle. What leaders can provide is clarity around the outcome, the principles that will not be compromised and how each person contributes to the broader mission. Integrity, excellence and collaboration become especially important when transformation crosses technology, business, product, operations, security and data. I also believe leadership will become more human as AI becomes more capable. Judgment, trust, curiosity and the ability to create clarity from ambiguity cannot simply be automated. Perhaps that is why teaching has always appealed to me. I enjoy taking something complex, finding the underlying pattern and making it understandable. The role of a technology leader is not to predict every change. It is to build an organisation that can absorb change faster than its competitors.