OskiBot

Inspiration

The inspiration for this project stemmed from the need for a streamlined, efficient way to access up-to-date information about UC Berkeley. With the vast amount of data available and the fast-paced changes on campus, students, faculty, and visitors often find it challenging to stay informed. We envisioned a solution that could provide instant, accurate information to enhance the UC Berkeley experience for everyone.

What it does

OskiBot is a one-stop chatbot designed to deliver the most current information about UC Berkeley. It provides users with the latest news, event schedules, academic updates, and campus resources. Accessible anytime, anywhere, OskiBot ensures users stay informed and connected with the UC Berkeley community effortlessly.

How we built it

To build OskiBot, we started by web scraping the UC Berkeley website to gather a comprehensive knowledge base of the most relevant and up-to-date information. This data included news, event schedules, academic updates, and campus resources. We then utilized the Claude Sonnet model for training, leveraging its advanced natural language processing capabilities to ensure that OskiBot could understand and accurately respond to user queries. By combining these technologies, we created a robust, real-time information retrieval system that seamlessly integrates data from multiple sources and delivers a user-friendly experience.

Challenges we ran into

*Building OskiBot came with several challenges. First, web scraping the UC Berkeley website to compile a comprehensive and up-to-date knowledge base was technically complex

*Next, training the Claude Sonnet model to effectively understand and respond to various user queries was challenging.

Accomplishments that we're proud of

We are proud of several key accomplishments in the development of OskiBot. First, successfully web scraping and compiling a comprehensive knowledge base from the UC Berkeley website was a significant technical achievement, ensuring that users receive accurate and up-to-date information.

Second, training the Claude Sonnet model to handle the diverse and complex queries related to UC Berkeley was a major milestone. This involved fine-tuning the model to understand specific campus-related terminology and context, resulting in a highly responsive and accurate chatbot.

What's next for OskiBot

OskiBot will evolve to support multi-modal interactions, including voice commands and integration with smart devices, to provide seamless access to UC Berkeley information anytime, anywhere.