Fragmentation of talent data has emerged as a fundamental constraint on the ability of corporate HR departments to make strategic decisions for their employees and continues as organizations systematically underinvest in unified data architectures.
According to the 2025 MuleSoft Connectivity Benchmark Report, only 34% of organizations are delivering a unified user experience across channels, despite running an average of 897 applications.
This distortion is cascading, as organizations spend an average of $1,283 per employee per year on training, even though 92% of corporate learning programs fail to link costs to measurable outcomes, according to a 2024 study by the Human Resources Development Institute.
In a recent conversation on the AI in Business podcast, Emerj Editorial Director Matthew DeMello sat down with Raoul Monroig, Bristol-Myers Squibb’s Intercon Regional People Organization Vice President.
Monroig emphasized the role of AI in unifying fragmented HR data into actionable talent intelligence and the need for HR leaders to focus their skill-building efforts on capabilities that truly drive business value.
The following analysis of their conversation explores two key insights for HR leaders.
- Integrate your fragmented talent data ecosystem: By unifying distributed employee information across HR technologies to build a trusted, validated data model, you can eliminate reliance on self-assessed competencies and enable accurate talent mobility and development decisions.
- Emphasis on disciplined skills and measuring scientific results: Replace programs that develop 15-20 skills with programs that develop them simultaneously with focused investment in just 2-5 key capabilities. This is measured by whether it actually improves business performance and ROI.
Listen to the full episode below.
guest: Raoul Monroig, Bristol-Myers Squibb Regional Vice President of Human Resources for Intercon
Expertise: HR leadership, talent acquisition, change management, global talent development.
Easy recognition: Raul Monroig Ruiz is a renowned human resources leader at Bristol-Myers Squibb, driving talent strategy and organizational transformation across diverse regions including China, Southeast Asia, the Middle East, Africa and Latin America. He drives the evolution of HR through change management and positive reinforcement learning to significantly enhance employee engagement and operational excellence within a global pharmaceutical manufacturing and distribution network.
Integrate the fragmented talent data ecosystem
Corporate HR teams face fundamental problems that most AI tools cannot address. That is, the data used to make talent decisions is not reliable enough to support talent decisions. The problem is not a lack of data. That means the information HR teams rely on is fragmented and full of partial fiction.
Mr. Monroig accurately distills this predicament: “We’re working with some of the data that should be available to us, but it’s not.” In other words, HR teams work across a sprawling technology stack that includes Workday, dashboards, Eightfold, and dozens of specialized tools, but these systems exist in isolation.
“Data is scattered all over the place. Most of us don’t have a system that brings everything together and allows us to work in a common view..And I think that’s the basis for how we work on individual development and team development. ”
– Raoul Monroig, Bristol-Myers Squibb Regional Vice President of Human Resources for Intercon
Raul points out that quality issues run deeper than disconnected systems. When HR teams attempt to map skills, build a pipeline, or design a development program, they start with a compromised foundation: self-assessment. Employees who evaluate their skills face the same incentives that shape public occupational profiles: the tendency to exaggerate strengths and minimize gaps.
As Monroig points out, self-reported competency data is structurally unreliable because people generally rate themselves as more competent than their peers.
Skill-building programs built on self-reported data are effectively built on sand. This flaw is not only methodological. This reflects a deeper structural issue in how companies approach talent intelligence.
As organizations seek to implement AI-driven mobility and skill matching workflows, the lack of trust becomes acute. These systems inherit the weaknesses of their inputs. Having fragmented data scattered across a platform is more than a technical inconvenience. Rather, investing in integrated insights is a strategic mistake.
You can resolve this issue because an alternate data source already exists at:
- performance indicators
- peer feedback
- project results and
- behavioral signals
These sources contain accurate information about what people can actually do. HR platforms continue to treat self-assessed skills and data in siled systems of record as sufficient.
Monroig argues that vendors should be selected and configured for the specific task of integrating fragmented data, rather than relying on general-purpose, off-the-shelf AI solutions. Machine learning systems can ingest disparate data sources and create a unified employee profile that surfaces patterns invisible to humans across siled systems. Instead of waiting for full integration, organizations can use AI as an integration layer to transform scattered signals into coherent talent intelligence.
For HR departments, the ability to integrate data in this way becomes a template for the broader enterprise, with AI as the fundamental engine that makes fragmented systems coherent and actionable, Monroig notes.
Disciplined skills focus and scientific performance measurement
Monroig argues that HR leaders often fail not because of intent but because of overload. Organizations frequently launch a broad portfolio of skills initiatives, including, among others:
- AI literacy
- Leadership development
- human resources management
- digital fluency, and
- Traditional compliance program
There is no clear hierarchy of which features actually drive business value. Raul says the results are predictable. “We’re going to be trying to develop a skills-building system, but frankly it’s going to cost a lot of money and we don’t know if it’s going to happen.”
An even more serious problem is that HR departments cannot measure impact if the underlying data is fragmented. Without a unified view of employee performance, progress, and accomplishments, attribution becomes impossible.
If HR can’t connect skills investments to increased revenue, retention, customer satisfaction, or team performance, there’s no basis for prioritizing which competencies are important.
Monroig’s view is that this measurement gap is pushing HR toward inclusivity: building everything and optimizing nothing, because the data foundation cannot support disciplined choices.
He argues that when HR departments lack a unified and trusted data architecture, three outcomes occur:
- No source: HR departments don’t know which skills correlate with business outcomes.
- No priority: All skills are equally important, so all are funded.
- Lack of strategic focus: Skills portfolios expand, but the impact remains untested.
In Monroig’s framework, the root cause is architecture, not ambition. Fragmented data makes a focused skills strategy impossible.
The result is more than just wasted spending. HR organizations often build large development systems without validating whether the skills they target impact business outcomes. Basic questions are not tested.
- Will this skill impact my final performance?
- Will it drive revenue or retention?
- Will customer outcomes improve?
For most companies, these questions remain unanswered, and Monroig points out that the state of the pharmaceutical industry exacerbates this blind spot.
“We need to focus on two, three, maybe five skills that will build our business over the next few years. [We should] We have to work relentlessly and very disciplinedly to build those skills with the right people and at the right level…We want to build leadership capacity, we also want to build talent management capacity. ” he claims.
– Raúl Monroig, Intercon Regional Vice President, Human Resources Organization, Bristol-Myers Squibb
Monroig also points to deeper uncertainties. Until now, HR departments have not operated on a validated scientific understanding of skill acquisition and behavior change. Reward systems and development programs have been rolled out over decades without rigorous testing.
Looking ahead, Monroig identifies three capabilities as essential for organizations to advance their AI adoption.
- curiosity
- agility and
- Customer service spirit.
Curiosity, as he puts it, is the willingness to experiment, even if the outcome is not guaranteed. He describes a Latin American employee who built a ChatGPT agent without any formal AI training, reducing brand launch preparation from 3 weeks to 30 minutes.
The willingness to test, iterate, and adapt reflects agility, or the ability to adopt new tools as they emerge. And a customer service mindset ensures that any AI application is not new, but firmly rooted in business needs.
