CrazyCats

Inspiration

With flight prices constantly fluctuating, travelers often face difficulty in planning and budgeting their trips. We aimed to create a tool that helps users make more informed decisions by visualizing airfare trends, enabling them to book flights at the optimal time for maximum savings.

What it does

CrazyCats is a real-time flight price analysis and visualization tool that combines historical data with live API updates to provide actionable airfare insights. Using Microsoft Fabric, we automate data ingestion, machine learning model training, and analytics to display insights on a dynamic, real-time dashboard. Users can explore key insights such as the Top 5 lowest-priced routes, price breakdowns by airline and number of stops, and average price trends over time. CrazyCats empowers travelers with data-driven visualizations, enabling them to make well-informed booking decisions.

How we built it

We built CrazyCats using Microsoft Fabric's Real-Time Intelligence (RTI) for seamless data integration from multiple sources, enabling real-time decision-making through an interactive dashboard. The solution processes data in real-time with Dataflow Gen2 and event streams, performs analytics using KQL queries, and provides continuous monitoring via a live dashboard. We trained a machine learning model on historical flight price data, incorporating key features such as departure and arrival locations and booking times. For data preprocessing, feature engineering, and model training, we used PySpark and Pandas in Fabric Notebooks to ensure a seamless and efficient data pipeline.

Challenges we ran into

A key challenge we faced was navigating Microsoft Fabric's wide range of services. With many similar functionalities across the platform, it required careful evaluation and testing to select the best tools for each aspect of our application. Additionally, collecting and cleaning large volumes of flight price data to ensure accurate predictions proved to be time-consuming. Finally, selecting and fine-tuning the best machine learning algorithms to handle real-time price fluctuations required extensive experimentation.

Accomplishments that we're proud of

We’re proud of developing a fully functional prototype that visualizes flight price trends and patterns, helping users make more cost-effective booking decisions. By integrating real-time data updates through Fabric Pipelines, we ensured users receive the most up-to-date flight prices for their trips. Additionally, our machine learning model achieved an accuracy of 85.35%, providing valuable predictions as more real-time data flows into the system.

What we learned

Through this project, we gained a deeper understanding of the complexities behind flight pricing and the many factors that influence it. We also improved our skills in data preprocessing, machine learning model selection, and real-time data integration. Leveraging Microsoft Fabric’s capabilities allowed us to streamline the entire workflow—from data retrieval and processing to feeding insights into the real-time dashboard.

What's next for CrazyCats

Looking ahead, we plan to expand CrazyCats by integrating more flight data from various platforms and routes. We’ll enhance our prediction model with additional data sources like fuel prices and regional demand trends, and introduce sentiment analysis from Reddit posts to capture external news that might affect flight prices. Ultimately, we aim to make CrazyCats the go-to tool for travelers seeking the best flight deals worldwide.