Navigating Careers in the Age of AI

Inspiration and Dataset

We were inspired by how AI is changing jobs and salaries. We wanted to see which skills are most valuable and how AI might impact careers, using data to uncover clear insights. We utilized a free dataset from Kaggle that fulfills our goal of understanding this concept further. AI Job Impact & Salary Dataset (2015–2035)

What it does

Our project analyzes how skills and AI risk relate to salaries. It uses data and machine learning to predict trends and show insights through clear visualizations.

How we built it

We built the project using Python and tools like Pandas, scikit-learn, and Matplotlib. We cleaned and processed the data, engineered key features, and applied models like Linear Regression and Random Forests to generate predictions and visualizations.

Challenges we ran into

We faced challenges with messy data, selecting the right features, and getting accurate predictions. Balancing model complexity and interpretability while keeping visualizations clear was also tricky.

Accomplishments that we're proud of

We’re proud that our project uncovers clear insights from complex data, successfully predicts salary trends, and highlights how AI affects careers. Our visualizations make the results easy to understand and actionable.

What we learned

We learned how to clean and analyze real-world data, engineer meaningful features, and apply machine learning models effectively. We also gained insight into how skills, AI risk, and salaries are connected in the workforce.

What's next for AI and Jobs - Datathon 2026

We want to take all these learnings going forward and improve our accuracy. We also want to explore interactive dashboards that let users explore trends themselves, and incorporate new data sources to better understand emerging skills and automation risks.

Built With

csv
machine-learning
matplotlib
omni
pandas
python

Submitted to

Data Heist 2026

Created by

I analyzed how AI risk, skill demand, and salary interact using our dataset. I engineered a metric called risk-adjusted job openings to estimate jobs that will be unaffected and built visualizations and scatter plots to highlight high-demand, low-risk careers. I developed both linear regression and Random Forest models to quantify how AI risk, skills, and salary influence safe job availability. This was my first hands-on experience with feature engineering and ML, and it gave me valuable insights that I will take with me for the future.

Shamikta Suresh Raj
I worked on building a machine learning prediction model that compared and analyzed the impact on salary over the years. I have also done the feature engineering for that category (year vs salary) and made that part of the presentation. I initially hypothesized that over the years, there would be a major impact on the salaries of people in tech. But the result turned out to be different. There was a consistent growth in salary, but not a major trend. This was my first time working on a hackathon-type project. The time crunch was really hard to manage, but this was a very good experience. It was a great learning experience.

Varsshine Rajesh
I primarily worked on analyzing the relationship between AI risk and job demand using our dataset. I focused on feature engineering, where I created a metric called risk-adjusted job openings to estimate how many jobs remain “safe” after accounting for AI automation risk. I also built a bubble chart in Omni to highlight which careers are both high in demand and low in AI risk. I developed a simple linear regression model to examine how factors like AI risk, skill demand, and salary influence safe job availability. It was my first time working with both feature engineering and machine learning in this way, and it was a really valuable experience!

Chloe Espenilla
I started off by cleaning the dataset that we used our project. After that was cleaned and shared with the rest of the team, I used something called feature engineering (which was my first time) and collaborated with my teammate, focusing on transforming two different variables (AI Risk vs. Salary) to be ML ready. I then attempted to create a model and a prediction whether or not an individual's salary has a connection with the average risk score in a certain job and ai risk category using a linear regression (low, medium, high). However, I ran into a problem and with the help of my teammate, we shifted to a random forest model, including more variables. I also started the sides and organized them before the rest of my teammates came over to the presentation. This was my first time doing anything related to the data science field, an experience that I will always remember.

Jeanna Park

Updates

Shamikta Suresh Raj started this project — Apr 05, 2026 12:08 PM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.