Solution Journey: AI
You can translate the document:
Overview
This document covers how the Denodo Platform enables the implementation of trusted, scalable AI solutions.
It’s divided into the following sections:
- Guiding Questions: A self-assessment checklist to determine the fit of your planned AI solutions, including autonomous AI agents and intelligent applications
- What: Explains what Denodo offers with its logical data management approach, using Query-RAG/Deep Research, to provide fast, accurate, and explainable AI answers and to enable trusted reasoning and decision support grounded in unified, secure data
- Why: Explains the strategic context, mission (real-time, compliant insights), and benefits (unified access, accelerated development, trusted data)
- How: The practical guide on implementing the solution. This includes:
- Customer References: Highlights real-world use cases and benefits for customers like Festo, Perkins-Coie, and Alexforbes
- Main Denodo Features for this Solution: Lists critical components: The Denodo AI SDK (for RAG), Denodo Assistant, MCP Server (for agentic architectures), virtual integration (semantic layer), and fine-grained security
- Steps to implement: The phased journey starting with Denodo Assistant (quick wins), transitioning to the AI SDK, and enhancing the setup with Denodo DeepQuery and MCP Server
- Main Resources: A catalog of videos, manuals, and tutorials for the core components
- Success Services: Details how Customer Success managers and services assist in the AI journey (e.g., architectural design, use case discovery)
Many organizations assume that when agentic AI initiatives struggle, the issue lies with the models or prompt design. In practice, the root cause is usually the data foundation.
These aren’t model limitations, they’re data challenges. Building reliable, scalable agentic AI requires a foundation that delivers live, contextualized, and governed data from the start.
Denodo addresses this by providing real-time, governed access to enterprise-wide data for AI, generative AI, and agentic automation. The Denodo Platform enables accurate, context-aware insights across structured and unstructured data, helping organizations move faster while maintaining trust, security, and control. Beyond answering questions, Denodo supports AI systems that can understand current business conditions, reason over trusted data, and act within clearly defined boundaries.
This document serves two audiences:
- Decision Makers or new Denodo users who need to determine if Denodo is the right fit for an AI solution, and who want to understand the rationale and benefits:
- These individuals should consult the sections entitled Guiding Questions, What, and Why.
- Implementation Teams or current Denodo users who want to understand the AI Solution's implementation with Denodo:
- This group should refer to the How section.
Guiding Questions
Denodo offers a wide array of solutions, each tailored to different business needs. These questions below are designed to help you quickly ascertain if this particular solution aligns with your specific requirements, providing clarity on whether it's the right fit for your challenges and objectives:
- AI Capabilities and Data Access:
- Is your AI progress being stalled by slow data lake provisioning or the inability to access "organization-specific" answers?
- Do you need to utilize all enterprise data (including live/fragmented sources), rather than just what is sitting in a central data lakehouse or warehouse?
- Agentic Capabilities and Operations:
- Does your use case require AI that can execute tasks and coordinate across multiple systems, rather than just answering questions?
- Must your AI systems maintain constant awareness of live operational conditions to support decisions?
- Operationalization and Scalability:
- Are you looking to implement a consistent data foundation that enables you to deploy and integrate AI use cases at scale?
- Do you anticipate deploying AI agents that operate continuously and integrate into your daily processes with minimal human intervention?
- Data Governance and Security for AI:
- Do you struggle to provide your AI with secure, governed, anonymized data that respects user permissions and global data sovereignty?
- Must your AI operate within strict business guardrails, so all decisions are explainable and based on trusted data?
If you answered "yes" to at least three questions, this solution contains critical recommendations to maximize your chances for success. Let's now dive into the What, the Why, and more importantly, the How — the actionable steps to make this solution a success.
What
AI with Denodo delivers instant, trusted insights by combining a logical approach to data management with Query-RAG (retrieval augmented generation), and it enables you to go even deeper with Deep Research capabilities. It unifies scattered enterprise data without requiring replication, then feeds precise, real-time results into AI models. It does this speedily using governed, secured metadata from across your enterprise. The outcome: reliable situational awareness, trustworthy reasoning, and a resilient foundation for AI applications ranging from conversational analytics to autonomous agents that support or execute business workflows.
Why
This section establishes the foundational context for the solution. It delves into the broader business problem, market trend, or strategic opportunity that the solution addresses. The following subsections answers the question: "Why is this solution needed?"
Vision
Agentic AI requires continuously updated information, as stale data causes risk and erodes trust.
Common challenges include:
- Inaccurate or incomplete responses due to stale or fragmented data
- Misinterpretation caused by inconsistent business definitions across systems
- Exposure of sensitive information from delayed governance
- Fragile behavior when agents interact directly with raw data
A centralized, unified data access layer enables AI models using Query-RAG to instantly discover, govern, and consume all enterprise data — across cloud, on-premises, and SaaS systems — at its source. This logical layer provides current, context-rich, and holistic information, enabling accurate, explainable insights and continuous decision support without data consolidation or costly movement. This approach connects unique data assets with AI, accelerates decision-making, facilitates compliance, and builds a future-proof foundation for transformation, unlocking the full value from data without system restructuring.
Mission
Denodo achieves this vision by retrieving relevant, authoritative data in real time from its governed logical data management layer, enabling all accessed data to comply with enterprise policies and regulations. This foundation supports AI systems that must reason using enterprise data and operate safely within business guardrails. DeepQuery enables users to ask complex natural language questions that translate into reasoning and planning that will autonomously build optimized queries across these governed, distributed data sources. Together, they simplify secure access to diverse enterprise data, eliminate duplication, and deliver real-time, compliant insights that accelerate decision-making and drive trustworthy AI-powered business innovation.
Why Denodo
In most data systems, like data lakehouses and business applications, AI is limited to the data that lives within that system. This creates silos and prevents organizations from defining and implementing global AI strategies.
The Denodo Platform can be used by AI agents to provide a unified, secure, real-time data layer that overcomes common data challenges like volume, dispersion, and security. The Denodo Platform’s universal semantic layer provides AI applications with contextually accurate, trusted data, while Query-RAG simplifies the development of AI applications by streamlining connectivity to a variety of LLMs while handling data embedding and orchestration. By delivering live situational awareness, shared business context, and centralized governance, Denodo enables AI systems to operate with confidence even in complex, distributed environments. This separation enables AI systems to focus on reasoning and orchestration while Denodo handles connectivity, semantics, and governance, reducing coupling and improving production resilience. This accelerates AI development, improves the accuracy of AI models, and enables organizations to build secure, reliable, and context-aware AI solutions without complex data movement.
The Denodo Platform’s seamless integration with agentic architectures – through MCP plus native support for REST-APIs and natural language querying via DeepQuery – empowers organizations to leverage advanced analytics without disrupting existing infrastructure. This combination of speed, agility, security, and broad connectivity makes Denodo the ideal partner to realize a virtualized, AI-augmented data architecture — accelerating innovation while reducing cost and complexity.
Benefits
Denodo delivers significant benefits to customers that seek to become AI-ready:
- Unified, real-time data access: Unified, policy-aware access to live operational and analytical data provides the situational awareness required for autonomous decision support and action.
- Accelerated AI Development speeds up time-to-market for AI initiatives by providing ready-to-use patterns (RAG, DeepQuery, MCP) and seamless LLM integration.
- Denodo Enables the Rapid deployment of AI-driven applications, decision support tools, and conversational experiences.
- Enhanced Query Productivity boosts the efficiency of data teams by enabling natural language queries (language-to-SQL), reducing the need for specialized coding.
- Trusted, Secure, AI-Ready Data delivers reliable, governed data for AI, mitigating the risks of hallucinations, facilitating regulatory compliance and protecting sensitive information, increasing trust in AI outcomes.
How
This section serves as the practical, actionable guide to implementing the solution. The details and real-world examples illustrating how this solution can be effectively put into practice are thoroughly explained in the following dedicated sections.
Customer References
Customers that implemented their AI use cases with Denodo gained significant value.
Festo, a leading provider of factory automation and process automation solutions, developed AI-powered capabilities that enable employees to access enterprise knowledge through natural language while maintaining governance controls. The company achieved:
- Accelerated GenAI use case development: Festo enabled seamless access to both structured and unstructured data for GenAI models:
- Measuring Sales Performance: Understanding the proportion of sales from online versus other channels
- Project Management Search: Enabling efficient searches for project-related data, such as start dates, end dates, and project staff
- Sales KPI Reporting: Providing upper management with up-to-date sales KPIs
- Reduced Technical Complexity for Users: Employees without technical backgrounds can get information using natural language.
- Enhanced Metadata Quality: Improved metadata management, including translations and detailed descriptions, facilitated better semantic understanding, enabling more accurate, contextually relevant GenAI responses.
Find the detailed use case here: Festo’s Innovation Journey: Leveraging GenAI to Transform Industrial Automation
Perkins-Coie, a multinational law firm, implemented Denodo's AI and data governance capabilities. By using Denodo to control the delivery of sensitive data, Perkins-Coie increased the reach of its AI chatbot from a mere 5% to 100% of its enterprise.
In addition, the company:
- Increased Efficiency:
- Significantly reduced the average response time for data-related inquiries
- Gained real-time decision-making capabilities for non-technical staff
- Increased Revenue:
- Increased customer loyalty with faster, more accurate insights
- Identified new services for existing customers
- Reduced Costs:
- If Perkins-Coie Chat can eliminate the need for other applications, this may also reduce licensing costs and technical debt.
Find the case study video, here: How Perkins Coie Conquered Data Fragmentation, Empowered Business Users, and Improved Client Engagement
Alexforbes, a leading financial services company in South Africa, manages over a million pension fund members monthly. The company provides a wide range of financial services, including investment solutions, financial advising, and insurance. To meet growing data demands and enhance operational efficiency, Alexforbes implemented a GenAI-powered data solution, centralizing its data ecosystem and enabling advanced analytics for seamless engagement from members to customers.
Supported by Denodo, Alexforbes was able to:
- Strengthen Data Governance and Efficiency:
- Denodo eliminated manual exports and data silos with secure, metadata-driven queries.
- Enhance Data Accessibility:
- Denodo’s logical data management capabilities seamlessly integrated disparate data sources, creating a unified customer view.
- Natural language querying simplified data retrieval, making critical information accessible even to non-technical users.
- Provide a Persona-Driven Experience:
- Denodo enabled all levels of users to interact efficiently with data across the ecosystem in real time.
- Multilingual query execution improved scalability across the nine countries where Alexforbes operates.
Find the case study, here: Enabling Personalized Data Engagement: Accelerating Intelligence with a GenAI Powered Solution | Denodo
Main Denodo Features for this Solution
This section covers the Denodo Platform features that are most critical – and directly utilized – when implementing this particular solution:
- AI and DeepQuery Infrastructure Features: to accelerate the development of AI applications, Denodo offers capabilities that simplify the connection of AI models to enterprise data.
- Denodo AI Software Development Kit (SDK): The Denodo AI SDK works with the Denodo Platform to simplify and accelerate the development of AI-driven applications and data-aware agents grounded in enterprise data using RAG. The SDK simplifies the process of grounding AI with enterprise data using the retrieval-augmented generation (RAG) framework. RAG-enabled applications can combine the power of retrieval-based and generation-based models to provide more accurate, more contextually relevant responses. The AI SDK streamlines the development process with multiple configurable LLMs and vector databases, enabling you to quickly create AI applications.
- Use the Denodo AI SDK - User Manual for reference.
- Denodo Model Context Protocol (MCP) Server: The Denodo MCP Server is an interface that enables AI agents and external applications to interact with the Denodo Platform. It provides a standardized protocol that enables any client to connect to Denodo's data virtualization engine. It is offered in two different options:
- Denodo Platform MCP: enables clients to understand the data model and make queries by generating accurate SQL from natural language questions as an extension of foundational VDP API functions. Refer to the Denodo MCP Server - User Manual for reference.
- AI SDK MCP Server: provides a full agent implementation with end-to-end Query RAG flow to interact with the Denodo Platform using natural language. These tools leverage the AI SDK's internal engine to handle discovery and query generation automatically. Relevant information can be found in the documentation for the AI SDK MCP.
- Improved Developer Productivity Features: Denodo Assistant is an embedded AI capability within the Denodo Platform that enhances productivity and simplifies data management tasks. It leverages semantic metadata and query intelligence to directly assist users in the interface.
- Natural Language Querying: Enables conversational data access without VQL/SQL knowledge.
- Smart Recommendations: Suggests related datasets, joins, and optimization options.
- Proactive Guidance: Offers query autocompletion, performance tips, and contextual help.
- Support for Unstructured Data: The Denodo Platform supports the vector type and includes AI VQL functions for unstructured data. More information can be found here: Large Language Model Functions and Embedding Functions
These capabilities accelerate preparation of trusted data foundations for AI applications and agents.
Besides these two categories of features specifically designed to streamline and speed up the development of AI applications using the Denodo Platform, three main foundational Denodo capabilities are particularly relevant for the success of AI initiatives:
- Virtual integration of data sources into a centralized logical data model: The logical data model is supported by a governed, secure semantic layer, and is critical for facilitating the development of AI-based applications on top of Denodo, for several reasons: It provides unified access to data, it enhances the accuracy and contextualization of metadata, and it provides rapid, simple, and flexible access to data sources.
- Fine-grained data security management: The Denodo security framework is a robust set of capabilities within the Denodo Platform for end-to-end data security, and it can also be applied to any AI solution. Denodo's Global Security Policies (GSPs) are important for AI development using the Denodo AI SDK because they provide centralized, consistent security management across all data views and users. This prevents AI models from only accessing unauthorized data, and it protects sensitive information such as personally identifiable information (PII) and intellectual property.
- Natural language communication with corporate data, without the need to vectorize actual data. Denodo’s architecture enables text-to-SQL (using Query-RAG techniques) to any data source, regardless of format and location. This simplifies your GenAI architecture, reducing replication pipelines specific for GenAI use cases. It also provides better security than RAG-based systems that require datasets to first be vectorized before they can be part of a GenAI answer; vectorized data cannot be secured with the same degree of control.
These foundational capabilities enable AI systems to access the right data, at the right time, with appropriate context and controls.
Steps to Implement
This section covers the steps and decisions that need to be considered when implementing AI solutions with Denodo. Our customers usually go through different phases to implement AI:
- Start with Denodo Assistant
- This aims to deliver immediate value by enriching data and addressing early implementation barriers.
- LLM Decisions: Key choices must be made regarding the LLM, including its location, the specific model, and the platform to be utilized.
- Data Usage: You need to decide what data can be leveraged and what security restrictions must be applied.
- This step quickly establishes a strong foundation for deploying AI use cases by enriching metadata — such as generating view and column descriptions, adding semantic meaning, and applying classification. This transforms a tedious, manual task into an AI-augmented, automated process that dramatically improves the usability and discoverability of data across the entire platform.
- Transition to the AI SDK
- Building upon the initial setup, this phase utilizes the metadata enriched during Denodo Assistant usage.
- LLM decisions are simplified, as Denodo Assistant already provides some learnings.
- The evaluation and comparison of several LLMs may be required.
- Depending on data sharing restrictions, some organizations are only allowed to use on-premises LLMs.
- Debugging is possible with external tools such as Langfuse.
- Data usage: additional data restrictions need to be taken into account:
- Some organizations use different datasets: For example, Test data (such as POC data) vs. Production Data.
- Security requirements may include:
- How data is shared
- Whether or not anonymization is required
- Start with a small subset of data products with a purpose, validate the solution with human agents, and incrementally add new subsets of data products.
- Enhance the AI SDK Setup
- DeepQuery can be adopted for advanced reasoning.
- Knowing the limitations of a chosen LLM for these additional queries, you may select a new LLM.
- AI SDK MCP Server can be adopted for integration with agentic architectures and external AI systems. It provides tools to query data and metadata with simple natural questions.
- Start with a small subset of data products with a purpose, validate the solution with human agents, and incrementally add new subsets of data products.
- Implement MCP Server on top of the Denodo Platform (no AI SDK needed) using the Denodo Connects MCP component:
- Note: If you run MCP Server without the AI SDK, agents will still have full metadata access and the ability to query views with VQL. However, in this case the agent would not have the RAG and chat capabilities of the MCP Server that comes with the Denodo AI SDK.
Recommended Resources
In this section, we list resources that can help you understand, implement, and optimize your AI-based solution implemented with Denodo.
|
AI Frequent Questions |
|
|
Introduction |
|
|
Denodo Deep Query |
|
|
Introduction |
|
|
Solution Overview |
|
|
Implementation Resources |
|
|
Support for Unstructured Data |
|
|
Solution Overview |
|
|
Implementation Resources |
|
|
Denodo Relevant Capabilities for AI: Security and Data Protection |
|
|
Solution Overview |
|
|
Implementation Resources |
|
|
Denodo Relevant Capabilities for AI: The Semantic Layer |
|
|
Solution Overview |
|
|
Implementation Resources |
|
|
Denodo Success Services |
|
|
The information provided in the Denodo Knowledge Base is intended to assist our users in advanced uses of Denodo. Please note that the results from the application of processes and configurations detailed in these documents may vary depending on your specific environment. Use them at your own discretion.
For an official guide of supported features, please refer to the User Manuals. For questions on critical systems or complex environments we recommend you to contact your Denodo Customer Success Manager.
Recommended resources Recommendations generated by AI
Questions
