CIOs and business leaders know they’re sitting on a goldmine of business data. And while traditional tools such as business intelligence platforms and statistical analysis software can effectively surface insights from the collated data resources, doing so quickly, in real-time and at scale remains an unsolved challenge.
Enterprise AI, when deployed responsibly and at scale, can turn these bottlenecks into opportunities. Acting quickly on data, even ‘live’ (during a customer interaction, for example), is one of the technology’s abilities, as is scalability: AI can process large amounts of information from disparate sources almost as easily as it can summarize a one-page spreadsheet.
But deploying an AI solution in the modern enterprise isn’t simple. It takes structure, trust and the right talent. Along with the practical implementation challenges, using AI brings its own challenges, such as data governance, the need to impose guardrails on AI responses and training data, and persistent staffing issues.
We met with Rani Radhakrishnan, PwC Principal, Technology Managed Services – AI, Data Analytics and Insights, to talk candidly about what’s working — and what’s holding back CIOs in their AI journey. We spoke ahead of her speaking engagement at TechEx AI & Big Data Expo North America, June 4 and 5, at the Santa Clara Convention Center.
Rani is especially attuned to some of the governance, data privacy and sovereignty issues that face enterprises, having spent many years in her career working with numerous clients in the health sector — an area where issues like privacy, data oversight and above all data accuracy are make-or-break aspects of technology deployments.
“It’s not enough to just have a prompt engineer or a Python developer. … You still need the human in the loop to curate the right training data sets, review and address any bias in the outputs.” —Rani Radhakrishnan, PwC
From support to strategy: shifting expectations for AI
Rani said that there’s a growing enthusiasm from PwC’s clients for AI-powered managed services that can provide both business insights in every sector, and for the technology to be used more proactively, in so-called agentic roles where agents can independently act on data and user input; where autonomous AI agents can take action based on interactions with humans, access to data resources and automation.
For example, PwC’s agent OS is a modular AI platform that connects systems and scales intelligent agents into workflows, many times faster than traditional computing methods. It’s an example of how PwC responds to the demand for AI from its clients, many of whom see the potential of this new technology, but lack the in-house expertise and staff to act on their needs.
Depending on the sector of the organization, the interest in AI can come from many different places in the business. Proactive monitoring of physical or digital systems; predictive maintenance in manufacturing or engineering; or cost efficiencies won by automation in complex, customer-facing environments, are just a few examples.
But regardless of where AI can bring value, most companies don’t yet have in-house the range of skills and people necessary for effective AI deployment — or at least, deployments that achieve ROI and don’t come with significant risk.
“It’s not enough to just have a prompt engineer or a Python developer,” Rani said. “You’ve got to put all of these together in a very structured manner, and you still need the human in the loop to curate the right training data sets, review and address any bias in the outputs.”
Cleaning house: the data challenge behind AI
Rani says that effective AI implementations need a mix of technical skills — data engineering, data science, prompt engineering — in combination with an organization’s domain expertise. Internal domain expertise can define the right outcomes, and technical staff can cover the responsible AI practices, like data collation and governance, and confirm that AI systems work responsibly and within company guidelines.
“In order to get the most value out of AI, an organization has to get the underlying data right,” she said. “I don’t know of a single company that says its data is in great shape … you’ve got to get it into the right structure and normalize it properly so you can query, analyze, and annotate it and identify emerging trends.”
Part of the work enterprises have to put in for effective AI use is the observation for and correction of bias — in both output of AI systems and in the analysis of potential bias inherent in training and operational data.
It’s important that as part of the underlying architecture of AI systems, teams apply stringent data sanitization, normalization, and data annotation processes. The latter requires “a lot of human effort,” Rani said, and the skilled personnel required are among the new breed of data professionals that are beginning to emerge.
If data and personnel challenges can be overcome, then the feedback loop makes the possible outcomes from generative AI really valuable, Rani said. “Now you have an opportunity with AI prompts to go back and refine the answer that you get. And that’s what makes it so unique and so valuable because now you’re training the model to answer the questions the way you want them answered.”
For CIOs, the shift isn’t just about tech enablement. It’s about integrating AI into enterprise architecture, aligning with business strategy, and managing the governance risks that come with scale. CIOs are becoming AI stewards — architecting not just systems, but trust and transformation.
Conclusion
It’s only been a few years since AI emerged from its roots in academic computer science research, so it’s understandable that today’s enterprise organizations are, to a certain extent, feeling their way towards realizing AI’s potential.
But a new playbook is emerging — one that helps CIOs access the value held in their data reserves, in business strategy, operational improvement, customer-facing experiences and a dozen more areas of the business.
As a company that’s steeped in experience with clients large and small from all over the world, PwC is one of the leading choices that decision-makers turn to, to begin or rationalize and direct their existing AI journeys.
Explore how PwC is helping CIOs embed AI into core operations, and see Rani’s latest insights at the June TechEx AI & Big Data Expo North America.
(Image source: “Network Rack” by one individual is licensed under CC BY-SA 2.0.)
The post AI enables shift from enablement to strategic leadership appeared first on AI News.
