The SyncSphere Console is an AI-enhanced Human Resources platform for corporate networking and team bonding for remote or decentralized workforces. It leverages AI and Machine Learning to identify clusters of employees and recommend events and activities perfectly tailored to the employee preferences of that cluster. Whether you host a small business Happy Hour for 5 employees in a suburb or a Charity Bike Build for 50 employees scattered across a city, SyncSphere brings your employees together to offer the in-person experience anywhere.

The key data model that drives the SyncSphere console is what we call a “cluster”. The purpose of the cluster is to segment employees into regions based on their proximity to each other. We normalized location data into latitude/longitude coordinates, which we then used to run a density-based spatial clustering algorithm called DBSCAN to organize a sample of 300 employees in the Philadelphia and NYC areas into 14 clusters. Each cluster describes the employees that are within its range, the employees within them, their respective activity preferences, and a description of the center point, for example, a cluster of 25 employees with coordinates in the Manhattan area will be grouped into the “New York City 1” cluster. Once we’ve standardized all of the relevant activity preference and location data, we pull data on upcoming events that are within the bounds of the cluster using SerpAPI, an open source, free to use API that scrapes data from Google Events. The final piece of the puzzle is the machine learning model we used to rank upcoming events based on the preferences of employees within a cluster. We used an XGBoost ranking model trained on our preprocessed sample data to score each event based on relevancy. All of these attributes are contained within the cluster model, which drives the user interface that will provide HR professionals with insights to help make informed decisions on events and activities for employees in each sector.

Python, PyTorch