-- Swetha Talluri, data leadership expert and author of a paper published in IEEE Xplore proceedings, has published analysis and research addressing one of the most consistent challenges facing organisations investing in artificial intelligence: why AI initiatives stall at the point of scale, and what the data governance architecture required to prevent that looks like in practice.
Talluri's position, developed through both peer-reviewed research and enterprise transformation work, is reflected in a simple observation:
"AI doesn't fail in the model," Talluri says. "It fails in the data. When enterprises treat data as a product rather than a byproduct, AI finally becomes scalable."
A Research-Informed Framework Grounded in Real-World Patterns
Talluri's study, published in the 2026 IEEE International Conference on AI Engineering and Innovations proceedings, introduces a governance-aware AI framework that brings Master Data Management directly into predictive model pipelines. The framework is presented not as a universal solution but as a practical, evidence-based approach to reducing semantic drift and improving consistency in enterprise decision systems, particularly in environments where data definitions shift subtly across teams and systems.
Gartner research and industry guidance consistently identify poor data quality and weak governance as among the top barriers to AI scalability, patterns that Talluri's work addresses directly. Her study demonstrated measurable improvements in prediction consistency, reliability, and feature stability compared to metadata-driven and cloud-based MDM approaches, reinforcing the position that governance is a structural requirement for trustworthy AI rather than an optional enhancement.
Talluri also acknowledges the nuance that governance-heavy approaches can slow early experimentation if not rolled out iteratively, noting that balance rather than rigidity is key to sustainable implementation.
Impact Across Enterprise Environments
The principles reflected in Talluri's framework align with modernisation efforts in retail and supply chain organisations, where product and supplier data often exists across hundreds of systems. In one initiative she supported, centralising master data and applying governance standards reduced the time teams spent resolving conflicting records, accelerated updates across systems, and gave the organisation a single consistent view of its product information. With cleaner data flowing across platforms, analytics teams were able to shift their focus from correcting data to using it, a transition point many enterprises do not reach.
Why Enterprise AI Initiatives Stall
In an article published by the Boston Institute of Analytics, Talluri describes a pattern she observes consistently across organisations. Businesses rush into AI expecting transformation. Data foundations remain fragmented or inconsistent. Models produce unreliable insights. Teams lose trust in outputs. AI never scales beyond pilot stage.
McKinsey's global AI research reflects the same finding: organisations with strong data foundations are significantly more likely to achieve meaningful AI outcomes.
The Three Pillars of a Reliable AI Ecosystem
Talluri frames enterprise AI maturity around three interconnected layers. Data governance defines the policies, accountability structures, and quality standards that keep data trustworthy before it reaches a model. Master Data Management provides a unified and authoritative view of core business entities, and Talluri's research demonstrates that embedding MDM directly into AI pipelines strengthens consistency and reliability across decision-making systems. Integration frameworks connect data across platforms so organisations can analyse information holistically rather than through isolated silos. When these three layers operate together, AI becomes not just functional but dependable.
A Cultural Shift Behind the Technology
Talluri identifies data quality as an organisational challenge rather than a purely technical one. Successful data transformation requires collaboration across data engineering, analytics, product teams, and operational leadership. When organisations create shared accountability for data, they build cultures in which information is trusted, used widely, and ready to support AI-driven decision-making at scale.
About Swetha Talluri
Swetha Talluri is a data leadership expert with over 17 years of experience in enterprise data management, Master Data Management (MDM), analytics, and digital transformation. Her work spans large retail and supply chain environments, where she has contributed to data governance and data centralization initiatives that support scalable analytics and AI capabilities.
She is the author of research published in the 2026 IEEE International Conference on AI Engineering and Innovations proceedings through IEEE Xplore. She has contributed analysis on data leadership and enterprise AI to the Boston Institute of Analytics, and participates in professional communities including IEEE and Sigma Xi. Her work bridges academic research with real-world enterprise implementation, with a focus on building trusted data foundations for scalable AI systems. Connect with Swetha Talluri on LinkedIn at linkedin.com/in/swethatj.
Contact Info:
Name: Gianmarco
Email: Send Email
Organization: Xraised
Website: http://xraised.com
Release ID: 89194357
Google
RSS