Build your Data Agent

What’s a Data Agent

A Data Agent is an intelligent interface that allows to ingest and interact with various data sources such as PDFs, RSS feeds, or other structured content. Once created, a Data Agent can process this content and respond to queries, enabling natural language access to your data. allowing you to retrieve insights or answers dynamically.

Interface overview

The first step is to create your Data Agent, Once you have your AI Agent configured, you can list your new Data Agent with just a few simple steps.

1. Navigate to the My Data Agents

   1. Go to Deploy Data Agents

   2. Click on the Start Building to start the agent creation process.

   3. You’ll be redirected to the interface shown above, where multiple configuration options will be available

2. Upload a Display image

   1. Select a PNG or JPEG image that represents your data model.

   2. This image will be displayed in the Data Agents Marketplace to attract users.

3. Sync your Content:

   You can attach a PDF file, integrate an RSS feed, or connect an MCP URL each serving as a data source that defines the context and content available to your agent.

   1. PDF File Integration

      1. Train your AI agent by uploading one or more PDF files. These documents serve as structured sources of information that your agent can parse and learn from. The agent will extract and understand the content within the PDFs to provide more accurate and context-aware responses to user queries.

   2. RSS Feed Integration

      1. Enhance your AI agent by connecting it to one or more RSS feeds. These feeds act as dynamic sources of information, and your agent automatically extracts the latest content to stay up-to-date. By continuously parsing new entries, the agent improves its understanding and delivers accurate, context-aware responses in real time.

   3. MCP Client Integration

      1. You can connect your AI agent to an MCP URL, allowing it to access external data sources in real time. This integration enables your agent to fetch, parse, and utilize live data from connected services, expanding its capabilities beyond static files or feeds

         

         Pro Feature (Free for Now): While MCP integration is typically a premium feature, it is currently available for all users at no cost.

4. Add Basic Information:

   1. Name: Choose a name that clearly reflects your AI agent’s purpose or function. A descriptive name helps users easily identify and interact with your agent.

   2. Description: Provide a brief overview of your AI agent’s capabilities and functionalities. This description helps users quickly understand the agent’s purpose, the type of data it can access, and the tasks it can perform.

5. Add Instruction and Persona

   1. You can configure the agent’s response tonality and behavioural parameters, providing specific instructions that guide how it formulates and delivers responses, setting clear instructions and persona will ensure consistent and effective interactions with users.

      Here is an example: Your name is Nova, and you are a financial insights AI assistant with access to live market data. Your replies are professional, concise, and informative, but still friendly. Use charts, emojis, or bullet points when helpful. Respond directly to queries without unnecessary greetings or introductions.

The final step is to create the agent. Upon successful creation, it will be listed under the My Data Agents tab. The agent is initialised with private access, meaning only the owner can view or interact with it.

And that’s it!, You can always tweak your configurations and edit your content sources to further update your agent.