Death by Data?

It started innocently enough: A request for one simple dashboard. The team delivered a masterpiece—complete with drilldowns, trend lines and a predictive model that claimed to calculate the CEO’s mood. Soon, every department wanted their own version. HR added a maroon Mood Monitor. Legal wanted a chocolate-colored Compliance Chart. Logistics asked for an indigo Inventory Index. Sales sought a red Reality Response.

The result? By the next quarter, the dashboard had 14 widgets and required four servers to run. Meetings turned into interpretive art sessions: “If the blue line intersects the red line when indigo is high, what will be the CEO’s mood?” the CFO asked. “Is the chocolate-colored line inversely proportional to the maroon line’s rise?” Finally, the CEO asked: “To hell with all the lines. Can you tell me if we made money in the last quarter?” Silence from the team, the app, the cloud. Why? The server was busy calculating the impact of lunar phases on procurement.

If that anecdote made you smirk, these statistics should make you work: Forrester says 33% of data analysts waste more than 40% of their workday just cleaning data before it can be used for strategic decision-making. This creates a significant problem for employees because trying to pull out the most valuable data is akin throwing a net into the Great Pacific Garbage Patch and hoping to pull out a fish.

AI adds more complexity to the mix. “As more B2B buyers adopt GenAI tools to gather faster insights, marketing, sales and product leaders are facing intense pressure to integrate GenAI into their go-to-market applications to keep up,” Forrester reports. “Yet, 19% of buyers using AI apps feel less confident in their purchasing decisions due to inaccurate or unreliable information provided by GenAI.”

The worry? The explosion of new and untested GenAI functionality combined with lagging AI user skills could result in incidents that could lead to the loss of more than US$10 billion in enterprise value from declining stock prices, legal settlements and fines, Forrester adds.

The AI Paradox

There are no signs of AI slowing down. On the other hand, the global market for AI infrastructure is surging towards record levels, with spending projected to hit USD 758 billion by 2029. “In Q2 2025 alone, organisations boosted investments in compute and storage for AI deployments by 166% year-on-year, reaching USD 82 billion,” says Lidice Fernandez, an IDC group vice president.

IDC’s revised forecast reflects stronger-than-expected demand for GPU and other accelerator-based servers in the US, replacing earlier predictions of a slowdown in late 2025. AI investment surge will continue into 2026 driven by expanding pipelines from major vendors and buyers.

“Servers with an embedded accelerator are the preferred infrastructure for AI platforms,” Fernandez says. They account for 92% of the total server AI infrastructure spending and have grown 207% in Q2 2025. Accelerated servers will exceed 95% of the server AI infra spend by 2029, growing at an annual clip of 42%.”

What hath AI wrought? AI will indelibly alter how human workers perform work in the future. However, there will not be a jobs apocalypse. Instead, starting in 2028-2029, there will be jobs chaos created by the need to reconfigure, redesign, splinter and fuse more than 32 million jobs each year. That’s Gartner’s prediction.

“Every day 150,000 jobs will evolve through upskilling, while 70,000 more jobs will need to be rewritten, reworked and redesigned,” notes Helen Poitevin, a Gartner distinguished VP Analyst. “Leaders must plan their AI investments and goals to anticipate and manage these changes. They need to decide on their destination: to pursue human-first designs that emphasise supporting people in their work or select AI-first designs that aim to maximise efficiency by relying on AI to perform tasks.”

The goal is not a worker-free enterprise, but a work-redefined enterprise that is adaptive, creative and profoundly human at its core. Because being “AI-first” only succeeds when it is people-first. “The next era of enterprise performance will not hinge on the quantity of people employed, but on the quality of collaboration between humans and AI,” Poitevin says.

The paradox? Despite the explosion of new tools, most organisations have yet to see clear financial returns from AI. “You see AI everywhere except on the bottom line,” says Asutosh Padhi, McKinsey’s global leader of innovation and strategy. “The companies that will win the next decade aren’t the ones chasing the newest tools, but the ones turning innovation into sustained impact.”

The Data Paradox

Data adds another dimension to the discussion. According to a recent survey conducted by SoftServe, a US-based consultancy, 65% of corporate leaders fear that no one at their organisation fully understands if the data is aggregated or how to access it. Moreover, 58% say key business decisions are based on inaccurate or inconsistent data most of the time, raising concerns for companies across industries and borders.

“Many think it’s time to hit ‘refresh’ on their data strategy,” says Rodion Myronov, a SoftServe vice president. “As high as 73% think major updates or a complete overhaul is needed. Nearly all (98%) believe an updated data strategy would be required for strategic initiatives like Gen AI.”

An impactful data strategy is not about perfection, but prioritisation of the missing piece of the whole data puzzle, not tossing it aside for the next shiny new toy. “Establishing a mature data strategy can help reinforce organisational foundations for you to pursue bigger and better projects in the future,” Myronov adds.

That’s especially true with India where the huge market for products and services generates mountains of data. No wonder then that IT spending in India is set to cross US$176 billion in 2026, up 10.6% from 2025. That’s the latest forecast from Gartner. India’s data center systems segment will record the highest annual growth rate at 21% in 2026, continuing to outpace all other IT segments despite moderating from 29% in 2025.

“Data center systems spend is primarily driven by substantial AI infrastructure investments and multiple government programs aimed at strengthening the local AI ecosystem,” notes Naresh Singh, a Gartner senior director. “India has one of the largest consumer bases for AI services globally. It attracts international investment in local infrastructure to support this expanding user base. Evolving data privacy and sovereign cloud requirements will drive growth in this segment through 2026.”

So what’s the data paradox? Even though data has become ubiquitous, monetising it has remained elusive. Companies have invested heavily to unlock data’s latent value by building warehouses, creating dashboards, embedding analytics into their operations and setting up data product businesses.

These investments have delivered significant value, but far more value remains untapped. That’s because more than 90% of organisational data is unstructured, such as documents, images, social media posts and voice recordings. Unstructured data has remained out of reach for effective monetisation so far. GenAI changes that.

“New gen AI tools can clean, analyse and productise unstructured data quickly,” says McKinsey. “Critically, GenAI doesn’t make unstructured data usable—it makes it connectable—turning isolated signals into strategic intelligence. Advances in NLP allow companies to extract structured meaning from raw text, convert narratives, messages or transcripts into standardised, analysable formats.”

Leveling the Leverage

Data is the backbone but must be cleansed of duplicates, missing fields, incorrect and outdated entries. Once that’s done, how can organisations, including SMEs (small and medium enterprises) leverage the confluence of ABC (AI, big data, and cloud)? Here are half a dozen tips, in alphabetical order, for companies to consider:

• Activate Automation: AI embedded in cloud workflows (from MLOps to FinOps) could help automate data prep, anomaly detection and routine decisions. McKinsey’s 2025 survey shows 88% of organisations use AI in at least one function, with 64% reporting AI is enabling innovation—where high performers redesign workflows to capture automation benefits beyond pilots.
• Bespoke Behavior: Embedding AI on big data stacks could enable granular segmentation and offer real-time benefits. For example, AI-powered customer buying behavior could “anticipate” future needs. AI-driven marketing could deliver from 25% to 60% productivity gains and significant return on investment when properly scaled, reinforcing personalisation’s measurable lift.
• Choose Cybersecurity: Leading cloud platforms provide zero‑trust, encryption, identity controls and compliance tooling at scale, all critical as AI use expands. Misconfigurations and excessive permissions could drive many cloud breaches. Strengthen IAM (identity access management) and continuous posture monitoring while scaling AI solutions across cloud and CRM (customer relationship management).
• Deepen Democratisation: Self‑service analytics could bring trustworthy data to non‑technical customers. A 2025 study found 60% of leaders now prioritise democratisation, yet 55% of firms remain siloed with data scattered across the enterprise without a central repository for analytics and engagement. The most obvious answer: MDM (master data management) with real-time updated logs in centralised storage.
• Enable Elasticity: Cloud elasticity lets teams spin up compute and storage for AI training and big-data pipelines in minutes, then scale down when demand subsides. Capital expenditures of hyperscalers hit record highs in Q2 2025 with accelerated server spending up 76% and data center capex up 43%year-on-year to meet AI demand, shorten provisioning cycles and improving time-to-value for new workloads.
• Foster Functionality: Cloud-based AI and data platforms collapse geographic barriers, giving distributed teams secure, role-based access to the same datasets and models. Google Cloud’s 2025 trends highlight enterprise AI agents and assistive search could help employees retrieve multimodal knowledge across regions, boosting cross-functional collaboration and execution.

Businesses today have access to an unprecedented volume of data, leading to data abundance that can hinder rather than help. Decision-makers become so bogged down in analysing data that they are unable to decide. Employees spend excessive time searching for the right information. Analysts waste a sizable portion of their day just vetting data before it can be used. Could using AI to do the heavy lifting be one solution?

Since we started on one corporate fable, let’s end with another: When IT announced the seamless migration to the new analytics platform, everyone cheered. Finally, one source of truth. On day one, the sales department uploaded 2GB of spreadsheets. Marketing added dashboards with complex filters that required a PhD to sort. Finance uploaded PDFs without converting them to Excel.

By week one, the system crashed under the weight of duplicate records and emojis in column headers. “We need a cleanup!” the CTO said. That translated to interns manually deleting 27,000 rows of placeholder data. After a week, the platform displayed: Data integrity 99%. Except no one could now find last quarter’s revenue figures.

Raju Chellam is a former Editor of Dataquest and is currently based in Singapore, where he is the Editor-in-Chief of the AI Ethics & Governance Body of Knowledge, and Chair of Cloud & Data Standards.

maildqindia@cybermedia.co.in