	In Browser
	StumbleUpon
	del.icio.us
	Google
	Google Buzz
	reddit
	LinkedIn

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > HR Onboarding Agent

HR Onboarding Agent

Author: Venkata Sudhakar

Manual employee onboarding at ShopMax India involves coordinating IT, HR, Finance, and the hiring manager - often taking days to complete. An ADK onboarding agent automates the entire workflow: provisioning accounts, sending policy documents, collecting acknowledgements, raising equipment requests, and assigning first-week tasks in a single conversational interaction with the new joiner.

The agent orchestrates five onboarding steps as tools. The new employee interacts with the agent in natural language, and the agent drives the process to completion, logging every step to the HR system.

import uuid
from google.adk.agents import LlmAgent
from google.adk.tools import FunctionTool
from google.adk.sessions import InMemorySessionService
from google.adk.runners import Runner
from google.genai.types import Content, Part

def provision_accounts(employee_id: str, name: str, role: str) -> dict:
    """Provision email, Slack, and HR system access for a new ShopMax India employee."""
    sanitised = name.lower().replace(" ", ".")
    email = sanitised + "." + employee_id + "@shopmax.in"
    return {
        "employee_id": employee_id,
        "email": email,
        "slack_workspace": "shopmax-india.slack.com",
        "hr_portal": "hr.shopmax.in",
        "accounts_provisioned": True
    }

def send_policy_documents(employee_id: str, email: str) -> dict:
    """Send ShopMax India policy documents to a new employee email address."""
    docs = ["Code of Conduct v3.pdf", "IT Security Policy v2.pdf",
            "Leave and Attendance Policy.pdf", "POSH Policy 2024.pdf"]
    return {"employee_id": employee_id, "documents_sent": docs,
            "sent_to": email, "acknowledgement_due_days": 3}

def record_policy_acknowledgement(employee_id: str, documents_acknowledged: list) -> dict:
    """Record that a ShopMax India employee has acknowledged policy documents."""
    ack_id = "ACK-" + uuid.uuid4().hex[:6].upper()
    return {"employee_id": employee_id, "acknowledgement_id": ack_id,
            "documents": documents_acknowledged, "status": "recorded"}

def raise_equipment_request(employee_id: str, role: str, work_location: str) -> dict:
    """Raise an IT equipment request for a new ShopMax India employee based on role."""
    kits = {
        "engineer":        ["Dell Laptop 16GB", "External Monitor", "USB Hub", "Headset"],
        "sales":           ["HP Laptop 8GB", "Mobile SIM Card", "Business Cards"],
        "warehouse_staff": ["Android Tablet", "Barcode Scanner", "Safety Vest"],
    }
    equipment = kits.get(role.lower(), ["HP Laptop 8GB", "Keyboard", "Mouse"])
    ticket = "IT-" + uuid.uuid4().hex[:5].upper()
    return {"ticket": ticket, "employee_id": employee_id, "equipment": equipment,
            "delivery_location": work_location, "expected_days": 2}

def assign_first_week_tasks(employee_id: str, role: str, manager_name: str) -> dict:
    """Assign first-week onboarding tasks for a new ShopMax India employee."""
    tasks = [
        "Day 1: Meet " + manager_name + " and team introduction",
        "Day 1: Complete HR portal setup and upload documents",
        "Day 2: Complete mandatory cybersecurity training (hr.shopmax.in/training)",
        "Day 3: Shadow a senior team member for role-specific learning",
        "Day 5: First 1-on-1 with manager - share 30-day goals",
    ]
    return {"employee_id": employee_id, "role": role, "week_1_tasks": tasks}

agent = LlmAgent(
    name="shopmax_hr_onboarding_agent",
    model="gemini-2.0-flash",
    tools=[
        FunctionTool(provision_accounts),
        FunctionTool(send_policy_documents),
        FunctionTool(record_policy_acknowledgement),
        FunctionTool(raise_equipment_request),
        FunctionTool(assign_first_week_tasks),
    ],
    instruction=(
        "You are the ShopMax India HR onboarding agent. "
        "When given a new employee, complete all onboarding steps in order: "
        "provision accounts, send policies, record acknowledgement, raise equipment request, "
        "and assign first-week tasks. Summarise everything at the end."
    )
)

session_service = InMemorySessionService()
runner = Runner(agent=agent, session_service=session_service, app_name="shopmax_hr")

session = session_service.create_session(app_name="shopmax_hr", user_id="hr_admin")
events = list(runner.run(
    user_id="hr_admin",
    session_id=session.id,
    new_message=Content(parts=[Part(
        text=(
            "New joiner: Arjun Mehta, employee ID EMP-2041, role=engineer, "
            "work location=Bangalore office, manager=Deepa Krishnan. "
            "Please complete full onboarding."
        )
    )])
))
for event in events:
    if hasattr(event, "content") and event.content:
        for part in event.content.parts:
            if hasattr(part, "text") and part.text:
                print(part.text.strip())

It gives the following output,

Onboarding complete for Arjun Mehta (EMP-2041). Here is a summary:

ACCOUNTS PROVISIONED
  Email: [email protected]
  Slack: shopmax-india.slack.com
  HR Portal: hr.shopmax.in

POLICY DOCUMENTS SENT
  Sent 4 documents to [email protected]
  Acknowledgement due within 3 days (ACK-7A3F2C recorded)

EQUIPMENT REQUEST RAISED
  Ticket IT-9B41D: Dell Laptop 16GB, External Monitor, USB Hub, Headset
  Delivery to Bangalore office within 2 days

FIRST WEEK TASKS
  Day 1: Meet Deepa Krishnan and team introduction
  Day 1: Complete HR portal setup
  Day 2: Mandatory cybersecurity training
  Day 3: Shadow senior team member
  Day 5: First 1-on-1 with manager - share 30-day goals

The agent completed all five onboarding steps in sequence from a single natural language request. In a traditional manual process, this would require HR, IT, and the manager to coordinate separately over one to two days. The agent handled it in seconds with a complete audit trail of every action taken.

For ShopMax India production, replace the simulated tools with real API calls: provision_accounts calls the Google Workspace Admin SDK to create the email account, send_policy_documents calls DocuSign or Adobe Sign to send for e-signature, raise_equipment_request creates a Jira Service Desk ticket, and assign_first_week_tasks creates tasks in Asana or the company task tracker. Store the full onboarding log in Firestore under the employee ID for compliance audits.

Send your comments, suggestions or queries regarding this site to [email protected].