QCxC 2025

Inspiration

Our journey began with a shared curiosity and passion for data. As four Waterloo students from Quebec, we were inspired by the challenge of making sense of a dataset provided by Federato. We saw an opportunity to not only analyze raw data but also to transform it into actionable insights that could drive real business decisions.

What It Does

Our project is built around three core components:

Data Analysis: We processed and visualized the Federato dataset, using pie charts, bar graphs, and maps to reveal trends such as popular operating systems, user engagement metrics, and geographical distributions.
Machine Learning Model: We developed a model that predicts user behavior based on past interactions, enabling a more personalized experience.
AI Chatbot: We integrated a chatbot designed to answer Federato employees' questions swiftly. This tool helps employees extract key insights from the data without sifting through complex reports.

How We Built It

We leveraged a diverse stack of technologies:

Python: For processing data, building our analysis scripts, and creating the machine learning model.
Pandas, Matplotlib, Seaborn, and Numpy: To convert raw CSV data into structured dataframes, clean and analyze the data, and produce intuitive visualizations.
Tensorflow: For building and training our predictive model.
Flask API: To integrate the machine learning model into an accessible API.
React, Next JS, and Tailwind CSS: To design and build a responsive and engaging front-end interface.

By combining these tools, we successfully transformed a complex dataset into a comprehensive and interactive product.

Challenges We Ran Into

Our project was not without obstacles:

Data Processing: The dataset was enormous and fragmented, which forced us to split it into manageable CSV chunks. This iterative process of working on one chunk at a time meant reusing scripts across multiple datasets.
Understanding the Data: With numerous columns—some ambiguous and others vague—deciding which data to retain was a major challenge. Countless discussions and debates helped us distill the dataset to only the most useful information.
Creating the Heatmap: Mapping user interactions accurately required deep data insight and working with unfamiliar libraries like CartoPy. This step was crucial for visualizing geographic trends but demanded a steep learning curve.

Accomplishments That We're Proud Of

Data Insights: Our visualizations and charts now provide Federato with clear, actionable insights into user behavior and demographics.
Predictive Model: Our machine learning model can accurately predict user behavior, paving the way for a more personalized user experience.
AI Chatbot: The chatbot component significantly streamlines internal data queries, saving valuable time for Federato employees.

We are immensely proud of the tangible impact our work can have on Federato’s strategic decisions.

What We Learned

This project was a masterclass in data science and collaboration. We learned how to:

Tackle massive, unstructured datasets and transform them into actionable insights.
Leverage a broad array of technologies—from data cleaning and visualization to machine learning and front-end development.
Collaborate effectively under pressure, balancing creativity with structured problem-solving.

Our experience underscored the importance of perseverance, teamwork, and adaptability.

What's Next for QCxC 2025

Looking forward, we envision several exciting directions:

Expanded Analyses: With more time and resources, we aim to dive deeper into the data, providing even more comprehensive insights and a refined business plan for Federato.
Scalability: Exploring cloud platforms such as AWS or Azure could dramatically enhance our processing capabilities and support larger datasets.
Enhanced Chatbot: We plan to iterate on our AI chatbot, making it even more responsive and user-friendly for Federato employees.
Better API Inclusion: We plan itegrate our APIs better with the front-end.

We’re excited about the future and eager to see how our work can continue to evolve and drive meaningful change.

Thank you to the CxC organizing team, our sponsors (a special shoutout to Federato!) We hope you enjoy our product as much as we enjoyed creating it!