As AI continues to dominate the technology landscape, the data underlying the information these artificial intelligence solutions train on is under more scrutiny than ever.
According to Research Nester, the big data and business analytics market size is evaluated at more than $309.68 billion in 2025 and is projected to reach $940.44 billion by 2035. In 2026, the industry size of big data and business analytics is evaluated at $343.4 billion.
These trends are being fueled by the increasing demands of real-time data analytics, AI and machine learning integration, and flexible data storage and processing in the cloud, as well as other emerging architecture solutions.
Here, data professionals share their big data and database/infrastructure predictions for 2026:
In 2026, unstructured data will emerge as the backbone of AI innovation, redefining how enterprises harness intelligence across their organizations: As AI continues to advance, the availability of high-quality structured data is reaching its limits, creating what many analysts call a “data ceiling.” However, with an estimated 80–90% of enterprise data being unstructured, from documents and emails to images, videos, and design files, the potential to unlock its value has never been greater. This vast, often underutilized data holds the key to deeper insights, smarter automation, and more contextually aware AI systems. The next wave of AI progress will depend on how effectively organizations can access, govern, and activate their unstructured data. Doing so will require a strategic shift; one that prioritizes data quality, context, and security in equal measure. Unstructured data is the next iteration of data for AI. In 2026, having a comprehensive strategy for enterprise unstructured data is no longer considered “being a step ahead”, but vital for AI success.—Nick Burling, chief product officer, Nasuni
Enterprise AI infrastructure needs will change how data is stored, managed, used, and accessed by applications: Agentic AI and AI-driven workloads require reliable data storage, information streamed in real time, events organized by context , and the reuse of data for new models. Therefore, the infrastructure to support them will be expected to include powerful compute (CPUs, GPUs, TPUs), high-performance networking, scalable storage, and robust security and governance measures. That’s the reason why enterprise AI data infrastructure market is headed to a $7 trillion valuation in 2030. To meet the surging demand, companies like Dell are enhancing their AI data platforms now so enterprises can convert distributed data into more reliable AI outcomes in 2026—Tiago Azevedo, CIO at OutSystems
The enterprise data stack will become “agent-ready” by default: By the end of 2026, connectivity, governance, and context provisioning for AI agents will be built into every serious data platform. SQL and open protocols like MCP will sit side by side, allowing both humans and machines to query, act, and collaborate safely within the same governed data plane.—Redpanda’s CTO Tyler Akidau
Data operationalization sparks “digital twins” for data: For 20 years, data was something you looked at in a quarterly report. Now, it's being operationalized as a core asset of products themselves or may also be the product itself. However, the underlying data environment is a complex web of hidden dependencies that can be easily disrupted in many organizations. As they prioritize IT resilience, more companies in 2026 will invest in creating real-time “digital twins” of their data ecosystems—not to monitor the data itself, but to manage the complex “skeleton” of pipes that holds it all together.—Ram Chakravarti, CTO, BMC Software
A developer exodus is on the horizon: Some open source projects, of which PostgreSQL is one, are heavily dependent on "grey beard" developers, who are highly experienced in what they do. Younger developers on projects like these are far scarcer, often because the languages and processes used by the projects are not the modern ones they're used too. In PostgreSQL's case, a cultural shift could be made to use modern tools like GitHub Pull Requests rather than patches that are shared via email, but it can be hard to persuade the long-time developers to change their ways. An even large problem is the programming language in use. PostgreSQL is written in C, which few developers learn at school these days, being far more likely to be familiar with the likes of Python, JavaScript, or Golang. How can projects like PostgreSQL avoid becoming unmaintainable and obsolete as more and more of the experienced developers start to think about retirement?—Dave Page, vice president of engineering, pgEdge
A unified data estate becomes the strategic battleground: The era of focusing solely on GPU availability is coming to an end. The real competitive advantage lies in creating unified, global data estates that can power inference and generative AI at scale. Enterprises will realize that fast storage isn’t enough—orchestrating massive, decentralized, unstructured data into a single global namespace is now essential. In 2026, infrastructure players who can eliminate silos across sites, storage systems, and clouds will become the most strategic players in AI adoption.—Molly Presley, SVP for global marketing, Hammerspace
The new generation of cyber pros will finally fix data sprawl: In 2026, the biggest cyber threat will not come from a new malware strain or a state-sponsored attack. It will come from the growing sprawl of data. Sensitive information now flows through thousands of APIs, SaaS platforms, and partner ecosystems, often without clear ownership or enforceable controls. Even organizations with mature security programs struggle to track where their data actually lives. The result is a perimeter that no longer exists in any traditional sense. Every new integration expands the attack surface, and every handoff introduces another layer of risk. It’s a reality that keeps me up at night. The next generation of cybersecurity professionals may be the ones to change that. They bring a fresh perspective that is not tied to legacy thinking and are building adaptive, data-centric models for protection. In 2026, their creativity and pragmatism could be what finally restores control to an increasingly borderless world—Steve Cobb, CISO, SecurityScorecard
Return of the service mesh: Service meshes will make a strong comeback. Early excitement with the technology gave way to disillusionment because sidecar-based architectures were difficult to manage. The introduction of ambient mode has simplified adoption by moving proxies to the node level. This reduction in complexity will encourage renewed adoption; in 2026 Istio ambient mode will likely become the most widely used service mesh technology—Ratan Tipirneni, CEO at Tigera