Empowering Local Small Businesses with a Gen AI Review Assistant.

Nushanth yadav — Wed, 19 Aug 2026 13:34:21 +0000

In a rapidly growing city like Tirupati, local small businesses are adapting to online platforms. Walk into any neighborhood biryani restaurant or local motorcycle gear shop, and you'll meet hardworking owners who provide excellent service. However, online reputation management remains a major hurdle.

For my Google Gen AI Academy "Meet the Builders" project, I built a text-based AI solution to help local vendors communicate professionally with their customers.

The Local Problem

When a customer leaves a review on a delivery app or Google Maps, the business owner often struggles to reply. Language barriers or simply not knowing the "corporate PR tone" prevents them from resolving customer complaints—like a delayed order of Schezwan chicken fried rice—or thanking regulars properly.

The Solution: LocalBiz Voice

I designed a lightweight, text-only AI assistant using the Gemini 3.5 Flash API via Google AI Studio.

The concept is simple. The shop owner provides two pieces of text:

The customer's review.
A rough, casual thought on how they want to reply.

The AI takes those raw thoughts and transforms them into a polished, professional response in both English and Telugu.

Prototyping in Google AI Studio

Building this required zero complex coding upfront. I used Google AI Studio's Chat Prompt feature to set strict system instructions, ensuring the AI adopted a polite, empathetic persona suitable for customer service.

My System Instructions:

You are a professional public relations assistant for local small businesses. The user will provide a 'Customer Review' and their 'Rough Reply'. Generate a polite, professional response in English and Telugu.

Testing the Prompt:
I simulated a scenario where a customer complained about a late and overly spicy food order.

The shop owner's input: "sorry man, rain caused traffic. will make it less spicy next time."

Gemini instantly understood the intent and expanded it into a beautifully crafted apology that protects the restaurant's reputation.

*
*

The Code

Because this is entirely text-based, it is fast to run and uses the free tier. Here is the Python code generated by AI Studio:

import google.generativeai as genai

genai.configure(api_key="YOUR_API_KEY")
model = genai.GenerativeModel(model_name="gemini-3.5-flash",
                              system_instruction="You are a professional public relations assistant for local small businesses. The user will provide a 'Customer Review' and their 'Rough Reply'. Generate a polite, professional response in English and Telugu.")

user_input = """
Customer Review: "The schezwan chicken fried rice was way too spicy and the delivery took almost an hour."
Rough Reply: "sorry man, rain caused traffic. will make it less spicy next time."
"""

response = model.generate_content(user_input)
print(response.text)

The Impact