Career Canvas

screenshot of working chatbot with example prompt

Inspiration

A prevalent issue among students is their difficulty in selecting courses that match their interests and career aspirations. This challenge arises from the multitude of available courses, leading students to often be unaware of their relevance to their desired industry.

What it does

Career Canvas responds to user prompts based on FMs (foundation models) provided by AWS Bedrock to provide the customer (student) with course recommendations based on gathered data sources.

Data Sources:

UofT Academic Calendar (course description, pre-reqs/co-reqs, semester offered, professor etc.) -RateMyProf -customer input (profiles for each user stores data about what courses student has taken in the past, their likings, and how they performed on those courses)

How we built it

We used AWS web scraper to extract data from UofT Academic Calendar and RateMyProf and store in S3; used Q to create the the GenAI ChatBot using the AWS Workshop provided.

Challenges we ran into

Extracting data from RateMyProf was a challenging task since it was not in the acceptable format.

Accomplishments that we're proud of

A working demo of the GenAI chatbot that responds to user prompts based on uploaded data sources.

What we learned

What's next for GenCourse

take inputs from industry professionals about course relevance -performance of other students, review their profiles and analyze that data
incorporate user feedback to improve recommendations based on ML learning model developed using AWS SageMaker
gather more data about course like mark distributions, drop rates etc
create a timetable builder -compare between two programs -compare two courses based on different parameters like difficulty, content relevance to customer goals.