All of us have surely heard this quote by Tim Berners-Lee: “Data is a precious thing and will last longer than the systems themselves.”
As artificial intelligence (AI) accelerates the pace of enterprise transformation, the ability to harness data effectively has emerged as one of the defining leadership challenges of our time. At the forefront of this shift, Saket Saurabh, the CEO of Nexla, is helping organizations rethink how data can be transformed from a fragmented operational asset into a catalyst for innovation, agility, and consistent growth.
Under his leadership, Nexla has championed a modern approach to data integration and governance. This enables enterprises to unlock the full potential of AI with greater speed and confidence. Guided by a vision that combines technological sophistication with business relevance, he is shaping a future where trusted, accessible data becomes the foundation of intelligent decision-making. It is a future where organizations can build sustainable competitive advantage in an increasingly AI-driven world.
The Data Catalysts
Nexla is a leading AI-powered data integration platform that helps organizations that helps enterprises transform fragmented data into trusted, AI-ready assets. With more than 600 pre-built connectors and support for diverse integration methods, including ETL, ELT, streaming APIs, and agentic RAG, it enables organizations to simplify complex data operations and accelerate innovation at scale.
Trusted by global brands such as Autodesk, Carrier, DoorDash, Johnson & Johnson, LinkedIn, and LiveRamp, the team processes over one trillion records every month. Its consistent recognition in the Gartner® Magic Quadrant™ for Data Integration Tools and strong customer ratings reflect the company’s commitment to delivering dependable, scalable, and future-ready data solutions. As organizations deepen their investments in AI, Nexla continues to play a pivotal role in helping them turn data into a strategic advantage.
Infrastructure Advantage
Saket Saurabh has forged products and platforms across multiple technology cycles. During the processes he went through, he observed three recurring patterns in transformative technology that are often overlooked until much later:
1. Infrastructure Determines Long-term Value
While the application layer typically captures early attention, the infrastructure layer ultimately shapes success. In the mobile era, the spotlight was on apps, yet the greatest value accrued to platforms that solved connectivity, identity, and distribution. Similarly, in big data, a lasting advantage belonged to organizations that mastered data movement and reliability rather than analytics alone.
2. First-Principles Thinking Creates Differentiation
Saket Saurabh has observed that incumbents frequently attempt to adapt legacy paradigms to emerging technologies instead of reimagining them from the ground up. The organizations that consistently outperform their peers are those willing to challenge established assumptions and design for the future rather than retrofit the past.
3. Timing Shapes Competitive Advantage
Infrastructure is often perceived as a commodity until its strategic significance becomes undeniable. By that stage, the market is typically crowded. Those who recognize its importance early and invest with conviction are best positioned to build durable moats and sustain long-term leadership.
Production Reality
Saket Saurabh’s view is unequivocal when we asked him about the importance of data infrastructure and organizations overlooking critical challenges. He agrees with this, as he has observed that much of the enterprise AI conversation remains centered on model selection, prompt engineering, and application layer capabilities, while data infrastructure is frequently treated as a solved problem. In his assessment, this assumption fundamentally misreads the realities of enterprise AI.
Among the most consequential yet overlooked challenges is contextual quality.
He shares, “Organizations are spending heavily to connect AI to their data, but the question of whether that data is accurate, current, governed, and meaningfully documented for an AI system to use correctly is being skipped.”
He describes this emerging risk as the creation of ‘context swamps’, the AI era counterpart to the data swamps that emerged when organizations assumed that accumulating more data would automatically translate into greater value.
He also highlights the widening divide between demonstration and production. While AI agents can deliver impressive results in controlled environments with curated datasets, deploying them reliably across enterprise-scale systems is far more complex. Fragmented data estates, operational inconsistencies, and layered permission structures introduce realities that are rarely visible during pilot phases. From his outlook, the relatively low rate at which AI initiatives successfully reach production reinforces a strategic reality: sustainable AI success is ultimately determined by the strength, quality, and readiness of the underlying data infrastructure.
The Metadata Imperative
Nexla’s early commitment to metadata-driven architecture and virtual data products was rooted in a fundamental belief: data derives value only when accompanied by meaningful context. From the outset, the company recognized that moving data between systems was never the ultimate objective. The real challenge was making data usable by ensuring it carried with it the schema, lineage, quality indicators, access controls, and documentation required for effective decision-making. This philosophy ultimately led to the creation of Nexsets, the organization’s virtual data products designed to package context and governance alongside data itself.
As the AI era has unfolded, that conviction has become increasingly relevant. The team has seemingly forever maintained that the quality of context determines the quality of outcomes.
He shares, “When you are training a model or grounding an AI agent, the quality of the context you provide is the single largest determinant of output quality.”
Whether training models or enabling AI agents, organizations depend on trusted, governed, and well-documented data to generate reliable results. What was once considered a data management best practice has evolved into a foundational requirement for enterprise AI.
Saket Saurabh believes the industry’s outlook must now extend beyond conventional data engineering. Increasingly, the challenge is one of context engineering. It ensures that the right context is built, maintained, and delivered to the right system at the right moment. This shift is influencing how Nexla approaches innovation, how success is measured, and how organizations can unlock greater value from their data in an AI-driven world.
Building Trusted Agents
As organizations accelerate their adoption of autonomous agents, Saket Saurabh emphasizes that prolonged success will depend less on the capabilities of the technology itself and more on the foundations that underpin it. He argues that enterprise-scale autonomy requires a deliberate approach to governance, accountability, and operational readiness. To that end, he points to three critical capabilities that organizations must establish before autonomous agents can be entrusted with business-critical responsibilities.
1. Governed Data Access:
He is fixated on the opinion that an agent can only be as reliable as the data on which it operates. When the underlying data is outdated, inconsistent, or lacks the context necessary for accurate interpretation, the quality of the agent’s outputs inevitably deteriorates. Before introducing autonomous agents into mission-critical environments, organizations must establish data products that are governed, continuously maintained, thoroughly documented, and supported by appropriate access controls.
2. Clearly Defined Action Boundaries:
There is a fundamental distinction between agents that generate insights and those empowered to take action. Whether creating tickets, updating records, or initiating transactions, each level of autonomy demands a corresponding governance framework. He observes that many organizations have yet to define these operational boundaries across their critical systems, creating unnecessary exposure to risk and unintended consequences.
3. Auditability by Design:
He believes organizations must be able to reconstruct precisely what an agent observed, how it reached a decision, and what actions it ultimately executed. Such transparency is indispensable in regulated industries and increasingly important across all enterprise environments. In his view, governance and auditability should not be retrofitted after deployment. They must be embedded into the underlying infrastructure from the outset to ensure trust, accountability, and sustainable adoption at scale.
Technology Signals
Saket Saurabh believes that some of the most consequential technology decisions are not about identifying what is new, but understanding what truly is transformative. In evaluating emerging trends, he relies on a framework that helps determine whether a development represents a fundamental shift in capability or simply a new expression of existing ideas.
The first is whether a trend fundamentally expands what is possible or merely repackages capabilities that already exist. Transformative technology shifts extend the boundaries of what organizations can achieve, creating entirely new opportunities and operating models. Hype cycles, by contrast, often rely on new terminology to describe familiar capabilities, generating excitement without fundamentally altering the underlying landscape.
The second question centers on the movement of complexity. Saket Saurabh believes that every technology capable of simplifying the user experience inevitably transfers complexity elsewhere within the ecosystem. The critical task is understanding where that complexity resides and who is responsible for managing it. If those answers are unclear, the promised simplification may be more perception than reality.
Applying this lens to AI agents, he views the ability to deploy software capable of reasoning across systems and executing actions as a genuinely transformative advancement. It meaningfully expands what organizations can accomplish.
He adds, “The complexity of connecting agents to real enterprise systems, governing what they can do, and ensuring their outputs are reliable did not disappear.”
Rather, it has shifted downstream to data and infrastructure teams. In his view, this is where some of the most consequential work in enterprise AI is taking place, because while user experiences may become simpler, the underlying complexity remains both significant and enduring.
The Power Subtraction
Saket Saurabh believes that the ability to manage complexity is one of the defining characteristics of effective leadership. In his view, the difference between leaders who create clarity and those who create additional complexity often comes down to a few fundamental principles. In his view, leaders who simplify complexity have a precise understanding of the problem they are trying to solve. Those who add complexity, by contrast, often lack a sufficiently defined outcome against which decisions can be evaluated. Features are introduced because they appear to represent progress, while integrations are pursued because they seem to expand capability. Yet without a clear objective, organizations often struggle to determine when enough has been achieved.
He is clear that effective simplification requires a willingness to remove, not just add. While most organizations naturally associate growth with expansion, true simplification often demands difficult choices, including retiring capabilities that required significant effort to build. In his view, the most effective leaders do not see subtraction as a setback. They see it as a strategic decision that creates greater focus, clarity, and impact.
Context Quality
Saket Saurabh believes organizations are beginning to face a challenge he describes as context bloat. Much like the data bloat of previous years, the assumption that more automatically creates value is proving misguided. Organizations are providing AI systems with more documents, more transcripts, more metadata, and more tool definitions, yet outcomes do not improve proportionally because more information is not necessarily the right information.
Furthermore, he adds that the organizations that develop strong context engineering capabilities will gain a meaningful advantage. The ability to identify outdated information, organize knowledge effectively, attach business meaning and metadata, and deliver context at the right moment will increasingly separate leaders from laggards. He believes this is not simply a productivity improvement; it is the difference between AI that performs reliably in production and AI that succeeds only in demonstrations.
From a governance perspective, Saket Saurabh argues that context engineering demands greater rigor around data ownership and quality. When an AI agent acts on incorrect context, the consequences are immediate and visible. As a result, organizations will need to place far greater emphasis on ensuring that the information guiding AI systems is accurate, trusted, and relevant.
Evolving Human Expertise
Saket Saurabh is of the mindset that the future is already taking its shape. He points to Nexla’s Express platform as an example. By allowing users to describe their needs in natural language, the platform can identify data, establish connections, apply transformations, and deliver a governed data product. What once required skilled engineering expertise and significant time can now be accomplished in minutes.
At the same time, he does not believe automation reduces the importance of human expertise. Instead, it changes where that expertise matters most.
Saket Saurabh shares, “The engineers who thrive in an automated data engineering environment will be the ones who understand context deeply: what data actually means in a business context, how it should be used, where it cannot be trusted, and what governance model is appropriate for different use cases.”
As routine work becomes automated, human judgment becomes increasingly valuable.
He also sees this evolution as a shift from data engineering to context engineering. The focus is no longer simply on moving and transforming data, but on ensuring AI systems have the right context to operate reliably. In his view, that requires a different and increasingly valuable form of expertise.
Enduring Principles
He points to one important shift: a greater appreciation for timing. Earlier in his career, he believed that being right about a technology was enough. Over time, he learned that being right too early can be just as challenging as being wrong. In his view, markets must be ready for what a company is building, and leadership requires the patience and focus to stay the course until timing catches up with conviction.
What has remained constant is a focus on solving the customer’s actual problem rather than pushing a predetermined solution. Saket Saurabh believes every technology cycle produces impressive products that fail to address a real business need clearly enough to earn a lasting place within an organization. The companies that endure are those that can clearly demonstrate the value they create and the problems they solve.
Another principle that has remained unchanged is the importance of hiring people who are comfortable with ambiguity. Saket views uncertainty as a natural part of building a company. In his experience, the individuals who can think clearly and make sound decisions without having all the answers are often the ones who help build the most resilient organizations.
Clarity Drives
Extracting actionable intelligence from data is a challenge. Among the technology, organizational, or leadership challenges, he chose the leadership one.
In his view, the tools to move, process, and analyze data at scale are already mature. Most organizations also understand what good data structures and governance models should look like. Yet the challenge persists because data rarely becomes a true leadership priority in practice.
Saket Saurabh observes that problems emerge when ownership is unclear. Data quality slips when no one is accountable for it. Silos remain when no one is empowered to challenge them. And failures in pipelines often go unnoticed when no one is responsible for the outcome they are meant to support. For him, these are not gaps in technology, but gaps in accountability and intent.
He believes technology can enable better data systems, but it cannot replace leadership clarity. Real progress happens only when organizations explicitly define ownership, set clear standards, and enforce responsibility for outcomes.
AI Foundations
As AI agents consume, generate, and act on data autonomously, Saket Saurabh believes the shift is ultimately about moving from human-centric systems to agent-centric ones.
In his view, most enterprise software was designed for people, built around screens, forms, and workflows that match human reading speeds and decision-making. AI agents operate very differently.
He shares, “They call tools programmatically, process context at machine speed, and need data products that are structured for agent consumption, not human consumption.”
Saket Saurabh believes this requires three key changes. Data must be exposed through governed APIs and tools rather than just dashboards, since agents act through functions, not reports. Data also needs clear, machine-readable context so its meaning, quality, and permitted use are unambiguous. Finally, governance and auditability must sit at the infrastructure level, not within individual applications, since agents will operate across systems.
He sees protocols like MCP as an important step toward standardizing how agents connect to enterprise systems. But in his view, the real foundation still lies in well-governed data products and context layers that ensure those connections are actually reliable.
Business Evolution
When asked about whether the greater opportunity lies beyond productivity gains in organizational redesign or new business models, Saket Saurabh believes productivity is only the most visible layer of value, and also the most crowded.
In his view, most software companies today are competing on the purpose of doing existing work faster. While that is useful for adoption, it does not reflect where the more meaningful shift is occurring.
He believes the real opportunity lies in how organizations make decisions. Today, most decisions are still made with fragmented or incomplete context. AI systems that can dependably surface the right information at the right time have the potential to improve both the speed and quality of those decisions in a compounding way. He sees this less as a productivity gain and more as a fundamental shift in how organizations operate.
He also believes that entirely new business models will emerge over time as AI agents begin executing end-to-end processes. In that world, companies will increasingly be able to deliver outcomes instead of tools.
He adds, “Organizations will be able to offer outcomes rather than tools, because the tool can now execute the work. That is a meaningful shift in how value is created and captured, but it requires the data and context infrastructure to be in place first.”
Enterprise Readiness
A defining aspect for the most successful AI-powered enterprises in the upcoming five years will be those that model sophistication. This will be guided by how organizations build their data and context foundations today. In his view, the companies that treat these foundations as strategic assets now, rather than after AI capabilities mature, will be the ones best positioned to lead.
He states, “The pattern I expect to see is a divergence between organizations that built governed, AI-ready data foundations and those that rushed to deploy AI on top of whatever data infrastructure they already had.”
Others will try to layer AI onto existing systems without meaningful change. Over time, he believes the difference will become clear: the first group will be able to move faster and with greater confidence, while the second will continue to face a familiar constraint: AI that works in controlled environments but breaks down in production.
For leaders, Saket points to a few priorities that matter right now: defining and enforcing data ownership, ensuring data is structured and governed for machine use, building auditability into systems from the start, and being intentional about where human decision-making ends and agent-driven execution begins.
In his view, organizations that get these fundamentals right will not just see better productivity. They will build capabilities that fundamentally set them apart from competitors over the long term.
Also Read:- Cio Times Magazine for more Information
