Agents

Dataworkz Agents are software systems that are powered by LLMs and Tools and provide reasoning based problem solving coupled with tool execution to solve complex problems. Dataworkz Agents are enterprise-grade systems with robust capabilities around data security and access control. An extensive tool set enables Agent and Tool builders to build, test, introspect and manage Agent operations. A Dataworkz Agent is described using a JSON which captures all the configuration declaratively. This agent configuration is given a unique id and is used when leveraging the Agent API to invoke the agent.

An example Agent configuration file can look like this -

{
  "id": "fa430445-3bbc-4b5d-9cf4-0fc73c5a5a77",
  "name": "BusinessInsightsAgent",
  "agentPersona": "You are a helpful AI assistant specializing in analyzing business data.",
  "scenarioSelectionFailureMessage": "I could not follow what you meant regarding business data analysis. Please rephrase and try again.",
  "scenarios": [
    {
      "name": "BusinessDataAnalysis",
      "description": "Analyzes structured or semi-structured business data including sales, marketing, support, and operations to generate insights and recommendations.",
      "examples": [
        "How were last quarter's sales?",
        "What are the marketing trends for this year?",
        "How can we improve customer support?",
        "What operational changes can increase our efficiency?"
      ],
      "toolRefs": [],
      "instructions": "",
      "planningInstructions": "",
      "config": {},
      "disabled": false
    },
    {
      "name": "Help",
      "description": "If the users asks for help or wants to know what the agent can do.",
      "examples": [
        "help me",
        "What can you do?",
        "What can I ask?",
        "How can you help me?"
      ],
      "toolRefs": [],
      "persona": "",
      "instructions": "You are an agent that can help with the following - \nAnalyzes structured or semi-structured business data including sales, marketing, support, and operations to generate insights and recommendations..\n Please respond to the user with a helpful response.",
      "planningInstructions": "You can directly invoke join to answer the user's query",
      "config": {},
      "disabled": false
    },
    {
      "name": "Greet",
      "description": "If the user greets us or says hello or something similar",
      "examples": [
        "Hey",
        "Good morning!"
      ],
      "toolRefs": [],
      "persona": "",
      "instructions": "Please respond with a helpful message and a cheerful, professional greeting",
      "planningInstructions": "",
      "config": {},
      "disabled": false
    },
    {
      "name": "Exit",
      "description": "If the user says goodbye or indicates that the conversation is over",
      "examples": [
        "Thank you!",
        "I am all set."
      ],
      "toolRefs": [],
      "persona": "",
      "instructions": "Please thank the user for their time and interest.",
      "planningInstructions": "",
      "config": {},
      "disabled": false
    }
  ],
  "scenariosCount": 4,
  "description": "An agent that analyzes business data and generates actionable insights.",
  "contextValues": [],
  "tools": []
}