AITech Interview with Alberto Pan, Chief Technology Officer, Denodo

Why It Matters

Live, governed data eliminates stale or risky AI outputs, boosting accuracy and regulatory compliance. This accelerates AI‑driven decision‑making while preserving enterprise trust.

Key Takeaways

• Live, governed data powers accurate GenAI answers
• DeepQuery enables multi‑step, cross‑system investigations
• Explainability meets AI Act compliance and trust
• Semantic layer essential; vector stores insufficient alone
• Open MCP standards foster interoperable AI agent ecosystems

Pulse Analysis

Enterprises are grappling with fragmented data landscapes, where silos and inconsistent governance hinder the reliability of generative AI. Denodo’s logical data layer abstracts disparate sources without moving data, applying a rich semantic model that injects business context into every query. By delivering fresh, governed information in real time, the platform addresses the core AI challenge of “knowledge freshness,” allowing models to reason on up‑to‑date facts rather than static snapshots. This foundation is critical as organizations scale AI initiatives that demand both speed and compliance.

DeepQuery extends that foundation by transforming GenAI from a simple question‑answer tool into a research‑grade engine. It orchestrates multi‑hop queries across finance, CRM, supply‑chain, and telemetry systems, automatically reconciling results and presenting a structured report with explicit lineage, source attribution, and policy masks. Such transparency satisfies emerging regulations like the EU AI Act and satisfies executive demand for auditability. The capability also reduces the need for ad‑hoc data extracts, cutting operational risk and accelerating insight delivery from days to minutes.

Beyond the immediate product, Denodo’s integration with its AI SDK and support for the Model Context Protocol (MCP) signals a shift toward open, interoperable AI ecosystems. Developers can embed DeepQuery into custom copilots, combine it with other agents, and maintain centralized security controls. Pan’s advice—invest in a robust semantic layer rather than merely expanding vector stores—highlights a strategic imperative: metadata and context are the true differentiators for trustworthy, enterprise‑grade AI. Companies that adopt this approach can expect faster time‑to‑value, stronger governance, and a competitive edge in the AI‑driven market.