Career Granny

Inspiration & The Problem

Career Granny fills a crucial gap in education by addressing the often-overlooked area of counseling. With the rapid growth of the computer science industry, students are overwhelmed by the vast array of career choices and the skills required. However, over 50% of American school counselors and teachers lack the specific knowledge to guide students in this field. Career Granny aims to empower students by helping them navigate this complex landscape and make informed career decisions for free.

What does Career Granny do?

Career Granny helps novice users discover potential careers by entering a simple keyword or career name. It suggests a career, identifies the five most important skills and proficiencies they need, and provides learning methods along with free resources for each skill. Users can explore hundreds of career options and select one that aligns with their interests and strengths. For students who already know their strongest skills, Career Granny predicts a suitable future career and offers additional resources to further develop the required skills.

How does Career Granny address the prompt

Though there are millions of free resources on the web, Career Granny is a one-stop shop that compiles the most necessary information about hundreds of computer science careers and skills onto one page. This makes it much more efficient to find learning resources and advice for counselors and students who are overwhelmed with thousands of resources on the internet. Through Career Granny, students and their academic leaders will have the power of the internet compiled into a single website so that they can sharpen their skills and improve their chances in the competitive tech job market.

Technical Details

Career Granny was built using synthetically generated data, based on a large list of computer science careers and related skills. A dataset was created with proficiencies for various skills as columns and the corresponding career as the last column. A decision tree model was trained on this data, achieving about 85% accuracy. The entire project is coded in Python, using Streamlit for the interface, and is deployed on Azure for easy browser access.