EmotoMetrics

Inspiration

Many employers face the challenge in understanding their employee's overall satisfaction working within a team or department and therefore, feedback methods are often used to determine how the company can improve. There are several employees at large companies like ADP, which means that there is a high volume of employee feedback to parse through. We wanted to provide a solution to get the overall sentiment of employee feedback.

What it does

EmotoMetrics is an AI-powered employee sentiment analyzer designed to generate a one-word description about the emotion that is most likely correlated with an employee's feedback about a company. Currently, we are using a dataset or employee reviews that were generated by AI, but the intended purpose is to use feedback collected across ADP employees. By leveraging natural language processing, we were able to analyze the word choice and tone of an employee's textual review to generate a sentiment score and from there, generate a sentiment description. We hope that this sentiment analyzer is useful in making changes for growth at ADP within different departments and among diverse employees.

How we built it

We used Python as our primary programming language and the Google Cloud NLP API as well as Hugging Face to incorporate an AI component to analyze employee feedback.

Challenges we ran into

One challenge we faced was how we could incorporate the Google Cloud NLP API into Jupyter Notebook. We were familiar with how we could integrate the API on local computers and we used this background knowledge to research how we could integrate the API on Jupyter Notebook.

Accomplishments that we're proud of

We are proud of our ability to incorporate advanced NLP tools like Hugging Face to provide accurate sentiment analysis.

What we learned

We learned how to integrate cloud-based APIs and how to strategically parse through internal employee chats and feedback to accurately provide a emotion description.

What's next for EmotoMetrics

We plan to use more complex libraries to parse through large volumes of professional feedback as well as improve the EmotoMetrics dashboard to have greater usability features. Currently, a user has the ability to distinguish statistical employee sentiment by department/office for the temporary company.

Built With

• google-cloud-nlp-api
• hugging-face
• jupyterlab
• python
• vscode