Preetpal Singh, Group Managing Director at Xebia – Interview Series

Preetpal Singh is Group Managing Director at Xebia, where he leads global initiatives focused on building and modernizing digital platforms, enterprise applications, and data-driven systems. With more than 26 years of experience across financial services, healthcare, manufacturing, consumer, and high-tech industries, he has worked at the intersection of technology execution and business transformation. His career spans engineering delivery, go-to-market leadership, and large-scale operational change, with a consistent focus on turning emerging technologies into systems that operate reliably in production environments.

Xebia is a global digital engineering firm that designs, builds, and modernizes digital products and platforms for enterprises worldwide. The company combines software engineering, cloud and data modernization, Generative AI, and intelligent automation to help organizations develop scalable applications, modern architectures, and integrated AI capabilities. With teams operating across multiple regions, Xebia works with enterprises to embed new technologies into core systems and customer-facing platforms, ensuring that innovation is supported by strong engineering foundations.

You have spent more than 26 years leading technology, sales, and business transformation initiatives across multiple industries. How has that journey shaped how you approach innovation and growth today?

Over time, I’ve become much more focused on durability than novelty. Early in my career, I was drawn to the power of the technology itself. With experience, I began paying closer attention to what happens after the pilot phase, when systems have to integrate with legacy platforms, pass compliance reviews, and perform consistently under real workloads. Across industries, the turning point always seems to come when a promising initiative moves into production. Ownership becomes critical. Integration complexity surfaces. Risk tolerance is tested. That’s where transformation either matures or stalls.

Today, I approach innovation with a practical lens. At Xebia, we prioritize building and modernizing the platforms that sit at the core of operations. We think about long-term maintainability, auditability, and scalability from the start, because growth only sustains when the underlying systems are engineered to handle pressure over time.

Where do enterprises struggle most when executing on AI?

The challenges usually appear once AI moves beyond experimentation and into operational systems. Building a model is one milestone, but embedding it into underwriting workflows, ERP systems, or claims platforms introduces a different layer of complexity.

In one deployment scenario, a model had performed well in testing environments. When it was connected to a live transaction flow, questions around audit logging, override authority, and rollback mechanisms slowed the rollout. Compliance teams needed traceability. Risk teams required clarity on escalation paths. Those operational details ultimately determined the timeline.

AI initiatives gain traction when data pipelines are structured, production monitoring is in place, and accountability is clearly defined. Organizations that plan for those realities early tend to scale more smoothly.

How do you distinguish between automation that improves efficiency and transformation that reshapes a business?

Efficiency-focused automation often streamlines steps within an existing process. Broader transformation emerges when teams reconsider how the process itself is structured.

I’ve seen cases where organizations were preparing to automate approval chains that spanned more than a dozen touchpoints. Once decision logic was redesigned and validation checkpoints were consolidated, the workflow became significantly simpler. The impact extended beyond cycle time; it influenced accountability and risk distribution as well.

When you simplify the decision flow with the business first, the automation becomes easier and the gains stick.

What distinguishes transformation in highly regulated industries?

Highly regulated sectors tend to embed governance considerations directly into system design. In industries like financial services and healthcare, traceability and explainability are requirements before deployment, not after.

I’ve worked with teams that incorporated model validation gates into CI/CD pipelines and established detailed logging standards from the outset. Escalation paths and monitoring frameworks were defined early in the architecture discussions. That discipline shapes more resilient systems and prepares organizations for increasing autonomy in AI-driven environments.

How do you ensure AI investments with measurable business outcomes?

I encourage leadership teams to define the operational shift they expect to see. That could mean reducing onboarding time, increasing service capacity without proportional headcount growth, or lowering error rates in claims processing.

Instrumentation plays a critical role. The platforms we deliver include tracking for throughput, latency, decision accuracy, and adoption rates. When those indicators are visible in production dashboards, business leaders can connect technology investments to operational performance.How do strategic partnerships accelerate AI at scale?

Enterprise AI initiatives span infrastructure, data platforms, application architecture, and security. Coordinated execution across those layers requires depth and alignment. Partnerships create momentum when they are directly connected to engineering capability and operational execution.

Our expanded collaboration with partners like Google Cloud, Anthropic and others includes the development of repeatable industry solutions and the establishment of a global Center of Excellence to support scalable delivery. Structures like these reduce integration friction and provide consistency as deployments expand across regions and business units.

What misconceptions do you see around AI readiness?

Many organizations view successful pilots as evidence of readiness. In practice, readiness shows up in structured data environments, stable integration layers, defined governance protocols, and clear accountability models.

Production stability and governance maturity shape long-term success as much as model performance. A model can perform well in testing, but once it’s operating at full transaction volume, latency, monitoring, and failure handling become critical. Teams need clear rollback procedures, drift detection, logging standards, and defined ownership once the model is live.

As AI systems transition from advisory roles to decision-making roles, clarity around overriding authority and escalation paths becomes essential. When those guardrails are built into the architecture early, scaling becomes far more manageable.

How should leaders think about balancing innovation and risk as AI becomes embedded in operations?

Responsible AI is built into the system from day one and maintained through active oversight. Clear criteria around where systems can act independently and where human oversight remains essential helps prevent ambiguity later.

Ongoing monitoring is equally important, as data patterns shift and regulatory landscapes evolve. AI systems require periodic recalibration and review to maintain alignment with business and compliance expectations.

How do you maintain alignment between engineering execution and executive strategy?

Alignment gets easier when people are honest about the trade-offs. If you want to move faster, there are technical implications. If you give systems more autonomy, you’ll need stronger monitoring. If you customize deeply, you’re making choices about how easy it will be to maintain later.

None of that is bad. It just needs to be explicit. When those realities are discussed upfront, ambition and discipline don’t compete with each other. Leaders can push forward, engineers can build responsibly, and everyone understands what they’re signing up for. That kind of transparency builds trust and keeps progress steady over time.

Where should enterprises focus their strategic investments over the next several years?

Adaptability will separate the leaders from the rest. That usually comes down to strong fundamentals: clean data, scalable cloud infrastructure, disciplined deployment, and clear visibility into how systems are performing.

Companies that invest in that foundation, and build the right ecosystem partnerships around it, can adjust as markets and regulations shift. In the long run, the advantage comes from building operations and architecture that can handle whatever wave comes next.

Thank you for the great interview, readers who wish to learn more should visit Xebia.