Skip to content

Commit 7d579d5

Browse files
author
RobuRishabh
committed
README_added
1 parent 64faaa2 commit 7d579d5

File tree

2 files changed

+214
-3
lines changed

2 files changed

+214
-3
lines changed

‎README.md‎

Lines changed: 212 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1 +1,212 @@
1-
# RAG-based-Flight-Query-Bot
1+
# RAG-based Flight Query Bot
2+
3+
## Project Overview
4+
The RAG-based Flight Query Bot is a Streamlit-powered web application that allows users to query flight information using natural language (e.g., "Show me flights from New York to London"). It uses Retrieval-Augmented Generation (RAG) with an Ollama language model to extract entities from queries and retrieve flight data from a mock database. The project is deployed on a local Kubernetes cluster using Minikube and includes a CI/CD pipeline with GitHub Actions for automated testing.
5+
6+
### Architecture
7+
- **Frontend**: Streamlit UI (`app.py`) for user interaction.
8+
- **Backend**:
9+
- `query_handler.py`: Processes queries and extracts entities using Ollama.
10+
- `ollama_api.py`: Integrates with the Ollama LLM for natural language responses.
11+
- `mock_database.py`: Provides mock flight data and search functionality.
12+
- **Deployment**: Kubernetes on Minikube with two services: `flight-assistant-service` (Streamlit) and `ollama-service` (Ollama server).
13+
- **CI/CD**: GitHub Actions runs unit tests on every push or pull request.
14+
15+
---
16+
17+
## Setup Instructions
18+
19+
### Prerequisites
20+
- **Windows, macOS, or Linux**
21+
- **Python 3.10 or 3.11**
22+
- **Docker Desktop** (for building images and running Minikube)
23+
- **Minikube** (for local Kubernetes)
24+
- **kubectl** (Kubernetes CLI)
25+
- **Git** (for cloning the repository)
26+
27+
### Step 1: Clone the Repository
28+
Clone this project to your local machine:
29+
```bash
30+
git clone https://github.com/RobuRishabh/RAG-based-Flight-Query-Bot.git
31+
cd RAG-based-Flight-Query-Bot
32+
```
33+
34+
### Step 2: Install Python Dependencies
35+
Create a virtual environment and install required Python packages:
36+
```bash
37+
# Create a virtual environment (optional but recommended)
38+
python -m venv flightquerybot
39+
40+
# Activate it (Windows)
41+
flightquerybot\Scripts\activate
42+
43+
# Activate it (macOS/Linux)
44+
source flightquerybot/bin/activate
45+
46+
# Install dependencies
47+
pip install -r requirements.txt
48+
```
49+
50+
### Step 3: Set Up and Run the Local Ollama Server
51+
The project uses a local Ollama server for language processing.
52+
53+
#### Install Ollama:
54+
- Download and install Ollama from [ollama.com](https://ollama.com).
55+
- Follow the installation instructions for your OS.
56+
57+
#### Pull the Model:
58+
```bash
59+
ollama pull llama2:latest
60+
```
61+
This downloads the `llama2:latest` model (default for this project).
62+
63+
#### Run the Ollama Server:
64+
```bash
65+
ollama serve
66+
```
67+
Keep this terminal open. The server runs at `http://localhost:11434`.
68+
69+
### Step 4: Set Up Minikube
70+
Minikube creates a local Kubernetes cluster.
71+
72+
#### Install Minikube:
73+
Follow the [official Minikube installation guide](https://minikube.sigs.k8s.io/docs/start/) for your OS.
74+
Example for Windows (using PowerShell as admin):
75+
#### Start Minikube:
76+
```bash
77+
minikube start
78+
```
79+
80+
#### Verify Minikube:
81+
```bash
82+
minikube status
83+
```
84+
Ensure `host`, `kubelet`, and `apiserver` are `Running`.
85+
86+
---
87+
88+
## Deployment Steps
89+
90+
### Step 1: Build the Docker Image
91+
Build the `flight-assistant` Docker image using Minikube’s Docker daemon:
92+
```bash
93+
# Set Minikube's Docker environment
94+
eval $(minikube docker-env)
95+
96+
# Build the image
97+
docker build -t flight-assistant:latest .
98+
```
99+
100+
### Step 2: Apply Kubernetes YAML Files
101+
Deploy the application and Ollama server to Minikube:
102+
```bash
103+
# Deploy Ollama server
104+
kubectl apply -f ollama-deployment.yaml
105+
kubectl apply -f ollama-service.yaml
106+
107+
# Deploy Flight Assistant app
108+
kubectl apply -f deployment.yaml
109+
kubectl apply -f service.yaml
110+
```
111+
112+
### Step 3: Expose and Access the Services
113+
#### Start Minikube Tunnel:
114+
```bash
115+
minikube tunnel
116+
```
117+
Keep this running to assign an external IP.
118+
119+
#### Get the Service URL:
120+
```bash
121+
minikube service flight-assistant-service --url
122+
```
123+
Example output: `http://192.168.49.2:8501`.
124+
125+
#### Access the App:
126+
Open the URL in your browser to interact with the chatbot.
127+
128+
---
129+
130+
## Usage
131+
132+
### Interacting with the Chatbot UI
133+
1. Open the Streamlit app in your browser.
134+
2. Enter a query in the text box and press Enter.
135+
3. The bot processes your query and displays flight details or a "no flights found" message.
136+
137+
#### Example Queries
138+
- `"Show me flight NY100"` → Shows details for flight `NY100`.
139+
- `"What are the flights from New York to London?"` → Lists matching flights.
140+
- `"Flights from Chicago"` → Shows flights departing from Chicago.
141+
- `"Flight XYZ999"` → Returns "No flights found" (invalid flight).
142+
143+
---
144+
145+
## CI/CD Explanation
146+
147+
### GitHub Actions Configuration
148+
This project uses GitHub Actions for Continuous Integration (CI) to ensure code quality. The workflow is defined in `.github/workflows/test.yml`.
149+
150+
#### Workflow Details
151+
- **Trigger**: Runs on every `push` or `pull_request` to the `main` branch.
152+
- **Environment**: Uses an `ubuntu-latest` runner with Python 3.10.
153+
- **Steps**:
154+
- Checkout code:
155+
```yaml
156+
- uses: actions/checkout@v4
157+
```
158+
- Set up Python:
159+
```yaml
160+
- uses: actions/setup-python@v5
161+
with:
162+
python-version: '3.10'
163+
```
164+
- Install dependencies:
165+
```yaml
166+
- run: |
167+
python -m pip install --upgrade pip
168+
pip install -r requirements.txt
169+
```
170+
- Run tests:
171+
```yaml
172+
- run: |
173+
python -m unittest discover -s tests -p "test_*.py"
174+
```
175+
176+
---
177+
178+
## Running Tests Locally
179+
To test before pushing:
180+
```bash
181+
# Activate virtual environment
182+
flightquerybot\Scripts\activate # Windows
183+
source flightquerybot/bin/activate # macOS/Linux
184+
185+
# Run all tests
186+
python -m unittest discover -s tests -p "test_*.py"
187+
```
188+
Expected output: `Ran X tests in Y.YYYs OK`.
189+
190+
---
191+
192+
## Documentation Notes
193+
194+
### Stopping Minikube
195+
```bash
196+
minikube stop
197+
```
198+
199+
### Cleanup (Reset Everything)
200+
```bash
201+
minikube delete
202+
```
203+
204+
### Troubleshooting
205+
If the UI doesn’t load, check pod logs:
206+
```bash
207+
kubectl logs -l app=flight-assistant
208+
kubectl logs -l app=ollama
209+
```
210+
Ensure `minikube tunnel` is running for external access.
211+
212+
---

‎mock_database.py‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -36,7 +36,7 @@ def search_flights(origin=None, destination=None, flight_number=None, airline=No
3636
# ✅ If flight number is provided, prioritize searching by flight number only
3737
if flight_number:
3838
matches = [flight for flight in flights if flight["flight_number"].lower() == flight_number.lower()]
39-
print(f"🔎 Flight number search results: {matches}")
39+
print(f"🔍 Flight number search results: {matches}")
4040
return matches
4141

4242
# ✅ If no flight number, apply standard search
@@ -57,7 +57,7 @@ def search_flights(origin=None, destination=None, flight_number=None, airline=No
5757
and (not airline or flight["airline"].lower() == airline)
5858
]
5959

60-
print(f"🔎 Found flights: {matches}")
60+
print(f"🔍 Found flights: {matches}")
6161
return matches
6262

6363
if __name__ == "__main__":

0 commit comments

Comments
 (0)