← Back to Projects

Employee Retention Analytics

MIT capstone project building predictive AI model for workforce retention

Role: Data Science Researcher Timeline: In Progress - Completion January 2026 Institution: MIT (Massachusetts Institute of Technology)

Overview

Currently developing a sophisticated AI-powered predictive analytics model as my capstone project for MIT's Applied AI and Data Science post-graduate program. This machine learning system will identify employees at risk of leaving organizations by analyzing historical data, engagement metrics, performance indicators, and various workforce factors. The model will integrate seamlessly with existing company HRIS systems to gather real-time data and provide actionable retention insights to HR teams and leadership.

Project Screenshot

The Industry Challenge & Solution

Companies across all industries face a common crisis: costly employee turnover that drains resources, disrupts operations, and undermines team morale. Most organizations only discover retention issues through exit interviews—far too late to intervene. The reactive approach of trying to retain employees after they've already decided to leave is expensive and rarely successful. HR teams need a proactive solution.

The predictive AI model I'm building will transform how companies approach employee retention by identifying at-risk employees months in advance, giving HR teams and managers time to implement meaningful interventions. By integrating seamlessly with existing HRIS systems and analyzing real-time workforce data through advanced machine learning and neural networks, this tool will provide actionable insights while maintaining employee privacy and building trust in AI-driven HR decision-making. This solution addresses the gap between reactive retention strategies and the proactive, data-driven approach that modern organizations need.

Technologies & AI Stack

Core AI Technologies

  • Python (Pandas, NumPy)
  • Llama Large Language Model
  • Neural Networks
  • Deep Learning Frameworks

Data Integration

  • HRIS System APIs
  • Real-Time Data Pipelines
  • SQL & Data Warehousing
  • Performance Management Systems

MIT Capstone Program

  • Applied AI & Data Science
  • Massachusetts Institute of Technology
  • Completion: January 2026
  • Industry-Focused Research

Project Components & Goals

AI-Powered Predictive Model

Developing neural network-based machine learning models that analyze 50+ variables including tenure, engagement scores, compensation data, promotion history, manager effectiveness, and performance metrics to predict employee flight risk with high accuracy.

LLM Integration with Llama

Leveraging Llama large language model to process unstructured data from employee surveys, performance reviews, and feedback, extracting sentiment and engagement signals that traditional analytics miss.

HRIS System Integration

Building seamless integrations with existing company HRIS platforms to automatically pull real-time employee data, ensuring the model has current information for accurate predictions without manual data entry.

Actionable Insights Dashboard

Creating intuitive dashboards that translate complex AI predictions into clear, actionable retention strategies for HR teams and managers, including personalized intervention recommendations and ROI tracking.

Expected Impact & Goals

Jan 2026

Project completion date

High Accuracy

Target prediction reliability

Real-Time

Live data integration

This MIT capstone project combines cutting-edge AI technology with deep HR expertise to create a practical, industry-ready solution for employee retention. Upon completion in January 2026, this predictive model will demonstrate how artificial intelligence and machine learning can transform reactive HR strategies into proactive, data-driven workforce management.