This project is a Streamlit + Machine Learning web app that helps employees and employers evaluate fair trade-offs between salary and performance from both directions.
- Employee perspective: salary recommendation for a target performance; expected performance for an offered salary.
- Employer perspective: salary strategy within budget (maximize performance / maximize ROI) and minimum salary to reach a target performance.
The application features an interactive Streamlit interface where users enter an employee profile (job title, education, work hours, projects, etc.) and receive AI-powered recommendations through trained machine learning models.
It frames two unhealthy extremes:
- A greedy employer often wants maximum performance for minimum pay.
- A greedy employee often wants maximum pay for minimum performance.
The goal is to discourage both by making the trade-off explicit and to encourage fairness. Each case exists to support a fair deal:
- A fair employee asks for a salary aligned with a target performance (Salary Recommendation).
- A fair employer sets performance expectations aligned with an offered salary (Performance Recommendation).
- A fair employer optimizes pay within a budget to maximize performance (Maximize Performance).
- A fair employer optimizes pay within a budget to maximize ROI (Maximize ROI).
- A fair employer finds the minimum pay required to achieve a target performance (Minimize Salary).
The optimizer enforces a legal minimum salary floor to prevent recommendations that enable underpayment or employer abuse. This floor is sourced from the OECD, which provides official, comparable minimum wage data.
At present, the optimizer is limited to U.S. minimum wage values. Incorrect data (e.g., wrong country, currency mismatch, or API failure triggering fallback values) could distort the floor and result in unfair or invalid salary recommendations.
root/
├── data/ # dataset + processed artifacts
├── models/ # trained model artifacts
├── notebooks/ # experiments + model development
├── preprocessors/ # encoders/scalers
├── scripts/ # optimization + helper logic used by app
├── app.py # Streamlit entrypoint
├── requirements.txt # pip dependencies
└── salary_performance_analyzer_env.yml # conda environment
To run the Streamlit application on your local machine:
- Navigate to the project root directory
- Run the Streamlit app:
streamlit run app.py
- The application will automatically open in your default web browser.
The application is also available online at: https://salary-performance-analyzer.streamlit.app
Employee Performance and Productivity Dataset from Kaggle
This project serves as a learning exercise in Streamlit and machine learning. The notebooks document the complete workflow from data preparation to model comparison and evaluation.
Copyright (c) 2026 La Wun Nannda.
Licensed under the PolyForm-Noncommercial-1.0.0. See LICENSE.