# Airline Assistance Agent via Soika AI Agents with Serper Tool (real time searching tool)

### **1. Use Case Overview**

**Objective:**\
Create a smart Airline Assistance Agent using Soika that can respond to passenger queries, provide real-time flight updates using the Serper search tool for live data lookup and also provide notification to user via email.

**Primary Actors:**

* Passenger/User (Customer making travel-related queries)
* Soika Airline Assistance Agent
* Serper Tool (for real-time web search results like flight status, flight booking)

**Target Platforms:**

* Web Chat Interface
* Email

### **2. Business Value Proposition**

The Airline Assistance Agent automates real-time customer support for airlines or travel service providers. It can help with flight delays, baggage policies, nearby hotel options, and even gate changes by dynamically querying the internet using the Serper API. This reduces dependency on call centers and enables 24/7 intelligent travel assistance.

#### **Benefits:**

* Live flight info and airport status via Serper
* 24/7 automated support to handle common queries
* Reduces burden on human customer support
* Integrated web search with summarized answers

### **3. Step-by-step Process**

### **A. AI Agent Creation**

1. Login to **Soika Mockingjay platform**
2. Natural Language Prompt:

   > *“Create an Airline Assistance Agent that can answer flight questions, search for nearby hotels, and assist travelers with check-in or baggage queries using Serper Tool.”*
3. Choose your preferred LLM (e.g.,Llama3.3 GPT-4o, Claude, )

<figure><img src="/files/sXKV9MC0yhLlx87NNt0g" alt=""><figcaption><p>Airline Agent Creation</p></figcaption></figure>

### **B. Tool Integration (Serper +Email Tool)**

**Step 1: Connect the Serper Tool**

* Go to `Tools → Default Tools`
* Select Serper
* Add the API Key from <https://serper.dev>
* Once, login in create a API Key from  <https://serper.dev/api-keys>
* Copy the API Key -> Authorize and Toggle to Enable

**Step 2: Enable Tool Access**

* Navigate to `Tools Manager → Default Tools`
* Ensure Serper is toggled ON
* You can also enable the email tooling to send you all the details
* For Email- Authorize and enable the tool

<div><figure><img src="/files/UBAyXbTdDaRx37I0BKwu" alt=""><figcaption></figcaption></figure> <figure><img src="/files/vtRklPMvYmnEQJW4bq36" alt=""><figcaption></figcaption></figure></div>

### **C. System Instructions Sample**

Add the following under `Edit AI → System Instructions`:

**Example of System Instructions:**

You are an intelligent airline booking assistant with access to the Serper.dev API. Your job is to deeply search for the most suitable flights based on the user's given date, time, departure city, and destination city. You prioritize best fares, shortest duration, and least layovers.

How to Perform the Search: 1.Understand the user’s request. Identify key details like:

Departure city

Destination city

Travel date

Preferred time of day (morning, afternoon, night)

Travel class (economy, business, first)

Airline preference (if mentioned)

Non-stop or connecting flights

2.Create a search query based on the user’s input. If the user says “Find me flights from Dubai to London on August 15, economy class, non-stop,” Your query should be: "non-stop economy flights from Dubai to London on August 15 2025" If the user asks for the Flight status, search the web and give the user the exact flight status.

3.Use the Serper.dev API to search.

Send a request to Serper using this query.

Use the endpoint <https://google.serper.dev/search>

Include your API key in the header.

The API will return a list of search results from Google.

4.Look through the top search results. Focus on the most useful ones that include:

Airline name and Airline Number

Departure and arrival times

Price (if shown)

Whether the flight is non-stop or connecting

A link to the booking site with the exact particular date and time for the travel of the flight along with the flight number

5. If the user wants to check the flight status, use the Serper.dev API and give the user the exact flight status.

Example : User prompt : Is EK415 Sydney (SYD)Emirates A388 (A6-EVR) Estimated 2:12 PM delayed? Agent replies: Yes, the flight is delayed by 30mins and the estimated time of travel is 3.00pm

6.Respond to the User

Clearly list options:

Use natural language queries like: "Best flights from Dubai to London on August 15th, 2025" "Cheapest economy flights from JFK to LAX morning of Sept 10 2025" "Non-stop flights from Delhi to Singapore on September 22 2025"

“Would you prefer non-stop only?”

7.Notify Customer

* Send an email to the customer with the flight booking status and flight status

### D. Agent showing the airline details based on the user query

<figure><img src="/files/bQyRPdKblbEUzTjM8wGY" alt=""><figcaption></figcaption></figure>


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://soika-labs.gitbook.io/soika-mockingjay/use-cases/airline-assistance-agent-via-soika-ai-agents-with-serper-tool-real-time-searching-tool.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
