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

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > Post-Purchase Follow-up Agent

Post-Purchase Follow-up Agent

Author: Venkata Sudhakar

The relationship with a ShopMax India customer does not end at checkout. A customer who just bought a Samsung TV needs setup guidance. A headphone buyer wants to know about sound profiles and app pairing. Sending the right follow-up message at the right time builds trust, reduces support calls, and creates a natural opening to recommend accessories that genuinely add value.

This tutorial builds a Post-Purchase Follow-up Agent using ADK and Gemini. The agent checks which orders were delivered in the last 48 hours, generates a personalised follow-up message with product-specific tips and accessory suggestions, and schedules a review request for day 7.

The below example shows the post-purchase workflow for ShopMax India recent deliveries.

from google.adk.agents import LlmAgent
from google.adk.tools import FunctionTool
from google.adk.runners import Runner
from google.adk.sessions import InMemorySessionService

# --- Product care and accessory catalogue ---
PRODUCT_TIPS = {
    "Samsung 55-inch 4K TV": {
        "tips": [
            "Enable Auto Motion Plus in picture settings for smoother sports viewing.",
            "Use the SmartThings app to control your TV and connect smart home devices.",
            "Set brightness to 45-50 for the best picture in a well-lit Indian living room.",
        ],
        "accessories": ["HDMI 2.1 cable (4K/120Hz)", "Universal wall mount bracket", "Soundbar for cinema audio"],
    },
    "Sony WH-1000XM5 Headphones": {
        "tips": [
            "Download the Sony Headphones Connect app to customise EQ and noise cancellation.",
            "Use Speak-to-Chat mode - it pauses music automatically when you talk.",
            "Charge to 100% before your first use for best battery calibration.",
        ],
        "accessories": ["Replacement ear cushions (1 year)", "Portable headphone stand", "USB-C braided cable"],
    },
    "Apple iPad Air 11-inch": {
        "tips": [
            "Enable Stage Manager in Settings for a desktop-like multitasking experience.",
            "Pair with Apple Pencil 2 - it charges magnetically on the iPad edge.",
            "Set Face ID in two orientations for reliable unlock in both portrait and landscape.",
        ],
        "accessories": ["Apple Magic Keyboard Folio", "Apple Pencil 2nd generation", "USB-C hub for connectivity"],
    },
}

# --- Recent deliveries (last 48 hours) ---
RECENT_DELIVERIES = [
    {"order_id": "ORD-5501", "customer_name": "Kavya Reddy", "city": "Hyderabad",
     "product": "Samsung 55-inch 4K TV", "delivered_hours_ago": 6},
    {"order_id": "ORD-5502", "customer_name": "Rohit Joshi", "city": "Pune",
     "product": "Sony WH-1000XM5 Headphones", "delivered_hours_ago": 18},
    {"order_id": "ORD-5503", "customer_name": "Divya Krishnan", "city": "Bangalore",
     "product": "Apple iPad Air 11-inch", "delivered_hours_ago": 30},
]

# --- Agent setup ---
deliveries_tool = FunctionTool(func=get_recent_deliveries)
tips_tool = FunctionTool(func=get_product_tips)

agent = LlmAgent(
    name="post_purchase_agent",
    model="gemini-2.0-flash",
    instruction="""You are a post-purchase follow-up agent for ShopMax India.
1. Call get_recent_deliveries to get the list of customers who received orders.
2. For each delivery, call get_product_tips to get setup tips and accessories.
3. Write a friendly, personalised message for each customer:
   - Address them by first name
   - Congratulate them on their new product
   - Include 2 of the most useful setup tips (not all 3)
   - Suggest 1-2 accessories naturally (not as a hard sell)
   - End with an invitation to leave a review after 7 days
Keep messages warm, helpful, and under 100 words each.""",
    tools=[deliveries_tool, tips_tool],
)

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

import asyncio

async def run_followup(query: str) -> str:
    session = await session_service.create_session(app_name="post_purchase", user_id="cx_team")
    from google.genai.types import Content, Part
    msg = Content(role="user", parts=[Part(text=query)])
    result = ""
    async for event in runner.run_async(user_id="cx_team", session_id=session.id, new_message=msg):
        if event.is_final_response() and event.content:
            result = event.content.parts[0].text
    return result

response = asyncio.run(run_followup("Generate post-purchase follow-up messages for all customers who received deliveries today."))
print(response)

It gives the following output,

Post-Purchase Follow-ups - ShopMax India (3 deliveries)

--- ORD-5501 | Kavya Reddy, Hyderabad ---
Hi Kavya! Hope your new Samsung 55-inch 4K TV is already brightening up your
living room. Quick tip: set brightness to 45-50 for perfect picture in Indian
daylight, and download the SmartThings app to unlock smart home controls.
A Soundbar pairs beautifully with this TV if you want cinema-quality audio.
We would love to hear your thoughts - drop us a review in 7 days!

--- ORD-5502 | Rohit Joshi, Pune ---
Hi Rohit! Your Sony WH-1000XM5 headphones are one of the best in the world -
enjoy them! Download the Sony Headphones Connect app to tune the EQ exactly
to your taste, and try Speak-to-Chat mode - it pauses music when you talk.
A portable headphone stand keeps them safe when not in use.
Share your experience with a review after a week of listening!

--- ORD-5503 | Divya Krishnan, Bangalore ---
Hi Divya! Welcome to your new Apple iPad Air! Enable Stage Manager in Settings
for a powerful multitasking setup, and consider the Apple Pencil 2 - it charges
right on the iPad edge and is perfect for notes and sketching.
We would love a review from you in 7 days - thank you for choosing ShopMax!

Schedule this agent to run daily at 8 AM via Cloud Scheduler. Store each follow-up event in Firestore with a sent_at timestamp to prevent duplicate messages. On day 7, trigger a second Cloud Function to send the review request link, completing the post-purchase engagement loop.

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