Unified Data Layer Speeds Trusted Enterprise AI adoption

Denodo is an international company dedicated to data management and integration, specializing in data virtualization and logical data management.

Q: How would you describe Denodo’s position in the data management and integration market in the Iberian Peninsula and Latin America? 

A: Denodo is the undisputed global leader in data virtualization and advanced data management, with a track record of more than 26 years in the market. More than just a basic integration tool, we offer a comprehensive data management platform that unifies local (on-premise) and cloud-based data sources. Our proposition focuses on providing a single access layer in a more agile and cost-effective manner than traditional alternatives, unifying a company’s information without the need to move it from its original sources. 

Our key differentiator lies in our ability to deploy a unified semantic layer that instantly translates the technical complexity of data sources into business language, eliminating the physical fragmentation of information. While traditional approaches require large-scale architectural overhauls that take years, we enable data for AI initiatives in just three months. This agility accelerates return on investment (ROI) and ensures robust, centralized data governance, safeguarding privacy, and access control without compromising daily operations. 

Q: What are the main challenges you are helping your clients solve, and how are you doing so? 

A: We help companies overcome the high failure rate in transitioning AI pilot projects to production environments, which is a critical challenge. AI engines and agents often face the enormous difficulty of querying multiple unstructured physical sources, which degrades technical performance and complicates security. We solve this by providing a unified logical avatar of all corporate information, enabling AI to extract value immediately without having to decipher the individual technical languages of each storage silo. 

We also mitigate security risks and AI model hallucinations caused by a lack of clear business context. We centralize operational governance through a single point of control that enforces advanced access policies, dynamic real-time masking of sensitive data, and anonymization of confidential information. By structuring data under a consistent business semantics framework, we guarantee the absolute accuracy of autonomous processes and protect the integrity of our clients’ operations. 

Q: What are the main industries you are working with, and what success stories could you share for each of them? 

A: We operate across industries with high operational demands, such as banking, technology, aviation, and strategic industrial sectors. In the global financial sector, a flagship case is JPMorgan Chase, where Denodo serves as the sole access layer for 50,000 users and manages more than 34,000 data products. This implementation generated a return on investment within six months, yielding recurring savings of over US$1 million annually. In the Iberian Peninsula, we optimize operations for institutions such as Mapfre, where we provide a real-time, interactive 360° view for over 7 million customers across all their sales channels. 

In the technological and operational sphere, we support Intel’s infrastructure with more than 200,000 concurrent users, 2,000 application programming interface (API) connections, and more than two million daily queries distributed across 300 data sources. Likewise, in the aviation sector, All Nippon Airways (ANA) processes more than 15,000 operational transactions per second on our virtualization architecture without experiencing any performance drops. 

In Mexico and Latin America, we collaborate with leading corporations such as Telcel, Claro Colombia, and the Tecnológico de Monterrey, while also driving innovation in the agricultural sector with Bayer Crop Science in Brazil. One standout project is a high-criticality security initiative for one of Mexico’s largest mining groups, spanning seven mining sites. In this project in progress, Denodo will process real-time data streams from biometric cameras to verify personnel identity, technical training, and the use of protective equipment before automatically authorizing physical access, demonstrating our resilience in complex industrial environments. 

Q: How does the Denodo platform address hybrid and multicloud complexity without compromising performance, security, or regulatory compliance? 

A: We address hybrid and multicloud complexity through a native connectivity architecture that spans from traditional sources to modern data storage repositories. By collecting metadata from more than 400 different sources, we build a unified logical avatar governed from a single centralized point. This enables the application of global compliance rules, dynamic masking of sensitive data, and restrictions based on the user’s Internet Protocol (IP) address, regardless of where the information is physically located. 

To mitigate the impact on technical performance when processing billions of records, we integrate a specialized AI engine that dynamically optimizes query strategies in real time. Additionally, our platform incorporates massively parallel processing (MPP) based on the open-source Velox engine, which intelligently delegates complex workloads to other components to streamline the cross-referencing of large volumes of data.

Finally, the system learns from the organization’s specific usage patterns during the first three months of deployment. Based on this continuous learning, the engine automatically recommends and implements data aggregations that are transparently persisted for dashboards and end users, ensuring a response speed superior to that of traditional infrastructures without requiring forced migrations. 

Q: The “AI Trust Gap Report” notes that 66% of organizations believe AI can only be trusted if it operates with near-real-time data. Is the Latin American market truly prepared to support AI architectures with such operational requirements? 

A: The Latin American market is in an unbeatable position to adopt these operational requirements, and, in the worst-case scenario, it is only three months away from being ready through the implementation of our platform. The need for real-time data is imperative for the transition to agent-based AI, where autonomous agents directly intervene in critical business processes, such as fraud assessment or insurance claims. Traditional architectures based exclusively on data duplication result in unsustainable latencies; we bridge this gap by connecting operating systems, making information from operational systems directly accessible for data analysis and ensuring reliable, up-to-the-second information in a secure environment. 

Q: Many companies in Mexico and Latin America continue to face challenges when it comes to turning AI projects into scalable and profitable initiatives. What architectural or integration mistakes most frequently limit the ROI of these investments? 

A: The most common mistake that limits return on investment is the maintenance of silos, both in terms of information and operational governance, within fragmented data architectures. When organizations attempt to scale AI pilot projects to production without a unified semantic layer, they replicate these silos in their agent-based AI predictive models, fragmenting security and regulatory compliance controls. Likewise, the absence of a unified business context amplifies the hallucinations of unsupervised algorithms, which jeopardizes the integrity of critical business processes and hinders the capture of real financial value. 

Q: To what extent does a company become vulnerable when it connects AI agents directly to critical systems without a robust governance strategy? 

A: Connecting AI agents directly to critical systems without a robust governance strategy poses an unacceptable risk that exposes highly sensitive information assets to uncontrolled environments. To mitigate this vulnerability, Denodo strictly acts as an intermediate security proxy; corporate data is never directly exposed to AI algorithms. Our platform centrally intercepts every query to authenticate the requester’s identity, evaluate the context of the request, and apply the necessary regulatory filters, eliminating any possibility of fraudulent access. 

Q: What are Denodo’s priorities regarding expansion and growth in Latin America? 

A: Our strategic growth priorities for the 2026–2027 period in Latin America are structured around three fundamental pillars. First, we are driving the adoption of data sharing in hybrid environments, incorporating generative AI capabilities so that business users can build complex data products using natural language. Second, we are establishing secure data enablement for AI initiatives as a business transformation imperative. 

Finally, we are accelerating our native integration with modern storage architectures (lakehouses) such as Snowflake, Databricks, and Microsoft Fabric, supporting advanced formats like Parquet, Iceberg, and Delta to ensure seamless technological interoperability without forcing regional companies to restructure their infrastructure.