MCP & AI agents
Our data product marketplace solution connects your AI agents with your data via a secure MCP server. This creates a platform for reliable, high-quality agentic AI that drives improved business performance.
Make your data marketplace a [ strategic driver ] of AI innovation
Transform your data marketplace to enable greater efficiency by connecting your AI agents through our MCP server, unlocking the full potential of your data for all your stakeholders — from data experts to business teams.
Give your teams a natural way
to query your data with an
integrated [ AI exploration agent ]
Maximize the use and impact of your data with Huwy, the AI agent designed for business teams. Covering every step of their journey through your data marketplace, Huwy lets users ask questions in natural language and receive reliable, contextualized answers in seconds—with no technical expertise required. What sets it apart? Connected to the Huwise MCP server, Huwy explores your actual data—not just the metadata—to accelerate access to insights and clearly multiply the impact of your data marketplace.
Expand the ways your teams can [ access your data ]
Boost data discovery and usage by making your data accessible directly from the tools your teams already use. With our MCP server, the AI environments deployed in your organization — as well as collaboration platforms like Slack or Microsoft Teams — can connect to your data marketplace, multiplying touchpoints and driving daily engagement with your data.
Identify the channels that [ maximize usage ] of your data
Benefit from complete observability of data usage by precisely identifying access channels — data marketplace, APIs, AI environments — to optimize data distribution, value creation, and impact throughout your organization.
Seamlessly connect your data marketplace to your chosen [ AI environments ]
Easily deploy your AI agents thanks to our MCP server’s proven compatibility with major AI environments — Claude Desktop, OpenAI, Mistral AI, Microsoft Copilot — while benefiting from a fully managed, secure service provided by Huwise.
Examples of [ AI agents ]
deployed by our clients
Huwy Exploration Agent
Huwy directly explores the full content of your data to deliver precise, actionable answers in seconds. It turns data searching into a conversation and drastically accelerates time-to-insight.
Quality assistant
This AI agent analyzes your data marketplace metadata and provides recommendations to optimize its quality.
Documentation assistant
This AI agent automatically generates clear descriptions of a dataset and its fields to enable easier understanding.
Insight assistant
This AI agent identifies and summarizes major trends in your datasets and usage, providing you with clear and relevant insights.
Business assistant
This AI agent combines local and national data to identify the best places to open new retail stores and business branches.
Give your AI agents
the [ best playground ]
to unlock their full potential
Boost the performance of your AI agents through a rich and structured environment, crafted by Huwise over the past 13 years. Originally designed to facilitate human access and data understanding, it now serves as an ideal environment for autonomous, efficient, and relevant AI agents. With over 50,000 public datasets, powerful APIs, and comprehensive documentation, your agents have all the tools they need to fully unlock the potential of your data.
Embrace an [ AI-ready data marketplace ] to harness the power of tomorrow’s agents
While our client’s initial AI use cases are already generating significant value, they represent just a fraction of the potential of a data marketplace connected to our MCP server. Our product roadmap aims to fully leverage agentic AI to automate marketplace management and maintenance tasks, allowing your teams to focus on areas where human expertise remains essential.
These agentic AI-based innovations already further simplify access for our clients to their data, and boost its value and impact. However, this is just the beginning of our agentic AI plans. We are looking closely at integrating a conversational agent to facilitate business users’ queries, as well as supporting data marketplace administrators by progressively enhancing our configuration and management features.
474% ROI in less than 6 months with Huwise's data product marketplace
Discover how Huwise helps business users spend half as much time searching for data, and data teams cut data preparation time in half - freeing everyone to focus on what matters: turning data into insights.
Learn morereduction in time spent searching for data
reduction in data preparation time
reduction in data duplicated work
per year saved for IT
FAQ
-
An MCP (Model Context Protocol) server is an open-source standard that acts as a universal connector between a data marketplace and AI applications, often described as the “USB-C of AI applications.”
It enables any AI agent or application—such as Claude, Mistral, OpenAI, or Copilot—to connect to a data source through a universal protocol, without requiring complex integrations.
In practical terms, it provides a new access channel to the data available in a data product marketplace, alongside API access and the marketplace interface. It makes data directly queryable using natural language through the enterprise channels users rely on most.
Huwise natively provides a fully managed and secure MCP server, operated by our teams and designed to meet the highest quality standards.
With the MCP server, AI agents can:
- Retrieve relevant data,
- Understand its structure and documentation,
- Use it to generate insights and recommendations or trigger automated actions.
-
The Huwise MCP server is designed to be compatible with the leading AI environments already used by organizations.
It is interoperable with the following AI environments:
- Claude Desktop (Anthropic)
- OpenAI
- Mistral AI
- Microsoft Copilot
It also integrates with collaboration platforms such as:
- Slack
- Microsoft Teams
Regardless of the organization’s chosen AI stack, teams can access data available in a Huwise data marketplace directly from their usual AI applications, without switching environments, signing in again, or waiting for a specific integration.
The MCP server is fully managed and secured by Huwise, removing the operational burden for IT and data teams while ensuring compliance with governance policies.
-
Huwise customers are already deploying different types of AI agents connected to their marketplace through the MCP server, covering both data management and business intelligence use cases.
Current examples include:
- Quality assistant: analyzes data product marketplace metadata and recommends improvements.
- Documentation assistant: automatically generates clear descriptions for datasets and their fields.
- Insights assistant: identifies and summarizes key trends based on datasets and observed usage.
- Business assistant: combines local and national data to support strategic decision-making, such as identifying the best locations for opening new retail outlets.
These agents operate autonomously by querying structured data from the data product marketplace to produce accurate, actionable results.
They free data teams from repetitive tasks while giving business users immediate access to data-driven insights.
-
AI agents are most effective when they have access to structured, rich, and well-documented data. This is precisely the type of environment Huwise has been developing for more than 15 years.
The Huwise data environment includes:
- More than 50,000 public datasets,
- Powerful APIs,
- Detailed documentation,
- Semantically optimized metadata,
- Data quality indicators.
These elements were originally designed to make data easier for humans to understand and use. Today, they provide an ideal foundation for autonomous AI agents.
When an agent connects to a Huwise data product marketplace through the MCP server, it can not only retrieve data but also understand its context, lineage, quality, and terms of use.
This enables more reliable and relevant AI outputs than when an agent operates on raw, undocumented data sources.
-
Security and governance are built into the architecture of the Huwise MCP server from the ground up.
AI agents connecting to the data product marketplace through the MCP server are subject to the same access control rules as users. They can only query the datasets and data products for which they have explicit permissions, while respecting:
- Security policies, down to the individual column level of a dataset,
- Restrictions based on user groups,
- Conditions defined in data contracts.
Organizations can therefore deploy AI agents with confidence, knowing that they cannot access sensitive data beyond their authorized scope.
The MCP server is fully managed by Huwise as an enterprise-grade secure service, avoiding the risks associated with poorly configured self-hosted connectors.
In addition, agent activity can be fully observed and monitored, enabling administrators to track the data channels used by AI environments and optimize how data is distributed.
-
In many organizations, one of the main barriers to data adoption is the distance between where data is stored (in a catalog or marketplace) and the tools employees use every day (messaging, AI assistants, and productivity tools).
By connecting the Huwise data product marketplace to AI environments and collaboration platforms such as Slack or Microsoft Teams through the MCP server, organizations bring data to users, rather than asking users to go and find the data themselves.
Business users can then ask questions in natural language and immediately receive data-driven answers directly within the tools they use every day. They no longer need to know:
- Where the data is stored,
- How to query an API,
- Or how to interpret a raw dataset.
This conversational, contextualized access model dramatically lowers the barrier to everyday data usage, increases the number of active users, and accelerates the cultural shift toward truly data-driven decision-making across the organization
