Employee Onboarding Agent

Agent Description:

The Employee Onboarding Agent assists new hires by answering HR policy-related queries and providing personalized status updates. It connects to an internal knowledge base to retrieve policy information and uses an API to fetch each employee’s onboarding progress, including pending tasks and next steps, delivering concise, contextual updates.

Purpose and Components
  • Purpose: To automate key aspects of the employee onboarding process, providing new hires with easy access to HR policies and personalized updates on their onboarding status and required actions.
  • Components:

    • Policy Retriever: An agent that searches an internal knowledge base (MCP) for relevant HR policies.

    • Status Fetcher: An agent that retrieves a specific employee's onboarding progress details from an API.

    • API Connector (GET): A tool used by the Status Fetcher to get employee onboarding data.

Supported Capabilities

  • Connecting to an internal HR Knowledge Base (MCP, likely SQL-based) to search for policy documents.

  • Retrieving document details like title, category, and summary based on employee queries.

  • Normalizing retrieved text (removing duplicates, filtering by version).

  • Performing a GET request using an employee_id to fetch onboarding details.

  • Extracting key status fields: onboarding_status, pending_documents, completed_modules, and next_action.

  • Summarizing pending actions and HR contact details for the employee.

  • Cross-checking pending documents against Knowledge Base categories.

  • Providing concise, human-readable update messages (e.g., “Your onboarding is 75% complete. Please submit your Health Insurance Form next.”).

LLM Used

  • GPT_4O_MINI

    Note: To learn more about the LLM and to modify its behavior, refer to the Configuring LLM settings section.

Sub-Agents

1. Policy Retrieval Agent

  • Role: Knowledge Reader

  • Scope: Internal HR MCP Knowledge Base

  • Description: This agent handles queries about HR policies. It connects to the internal knowledge base, searches for documents relevant to the employee's question, retrieves the latest version, cleans the text, and provides the necessary information (title, summary) to help answer the query or guide the employee.

  • LLM Used: Default (Inherits from parent agent).

  • Tools Used: None (implicitly connects to MCP Knowledge Base)

2. Onboarding Status Fetcher

  • Role: Status Tracker

  • Scope: Employee GET API (structured JSON data)

  • Description: This agent retrieves the specific onboarding progress for an employee. Using the employee's ID, it makes a GET request to fetch details like status, pending documents, and next steps. It then summarizes this information into a concise update message for the employee.

  • LLM Used: Default (Inherits from parent agent).

  • Tools Used: Request - Get

Tools Used:

  • Request - Get Tool: Fetches employee-specific onboarding status data from a remote JSON endpoint.

Note: For details on modifying the Tools, refer Tools Library section.
Agent Workflow Behavior Summary

  1. The agent likely starts based on an employee query or a scheduled check. The Policy Retrieval Agent (start node) is activated first.

  2. Based on the context (query or status check request), it either searches the knowledge base for policy info or triggers the Onboarding Status Fetcher.

  3. The Onboarding Status Fetcher uses the Request - Get tool to fetch the employee's current onboarding details using their employee_id.

  4. It extracts relevant fields (status, pending docs), summarizes the next required action, and generates a concise update message.

  5. The Onboarding Status Fetcher (end node) provides the final output, either the policy information or the status update message.

Sample Questions:

  • What is the company's policy on remote work?

  • Can you check the status of my onboarding?