Anxiety disorders are highly treatable, yet only 36.9 percent of those suffering receive treatment, according to the Anxiety and Depression Association of America. Tech company Ginger is trying to close that gap by providing patients with behavioral healthcare via text and video. NLP Data Scientist Setu Shah’s job? To develop text prompts for the virtual coaches using natural language processing.
It’s a lofty goal, and one that he’s using Python to accomplish.
“I’m using Python because I can go from thinking to prototyping solutions really quickly,” Shah said. “It employs simple syntax and enjoys broad community support for machine learning libraries.”
While Tempo Automation Product Manager Sherry Ren relies on Python for some of the same reasons, she’s using it for a very different purpose. Ren is improving data analytics models to plan and predict factory schedule and capacity at the electronics manufacturer.
“Python is great for data analysis and reliable for building the backbone of important processes.” Ren said.
Shah uses natural language processing to simplify emotional support coaches’ jobs. He takes advantage of beneficial tips and suggestions from coaches within Ginger’s text and video-based therapy and psychiatry archives to inform contextual decisions while the Python project is in development.
Tell us about a project you're currently working on in Python.
I’m working on a project to prompt text suggestions that our coaches can use in conversations with their members. I use natural language processing to understand the context of the conversation. I then train a model to leverage the context and predict a few appropriate coach response options.
The response options are populated from relevant responses coaches have actually used in the past, in similar contexts. I’m using Python because I can go from thinking to prototyping solutions really quickly. It employs simple syntax and enjoys broad community support for machine learning libraries.
I use natural language processing to understand the context of the conversation.’’
What impact will this project have on your customers, your company or the industry as a whole?
The difficulty in obtaining timely care is a pressing problem for the industry. This project will help our coaches respond to members faster when they encounter replies that are repetitive and recurring across conversations, leaving them more time to focus on the core problems that members face. For our company, this means the ability to serve more members.
Ren leads a data architecture project meant to provide quality, speed and transparency to Tempo Automation customers. The infrastructure will collect, manage and analyze the data generated during the printed circuit board (PCB) manufacturing process at the company’s factory. While this project will be internally beneficial, it also allows Ren’s team to create external tools that will increase productivity for users.
Tell us about a project you're currently working on in Python.
Our team is defining and building the architecture to collect data from multiple sources. We are applying extract, transform and load (ETL) processes and storing cleaned data in a data warehouse. Python is the best language for this use case because both software engineers and data scientists are very familiar with it. It’s also great for data analysis and reliable for building the backbone of important processes.
Our team is defining and building the architecture to collect data from multiple sources.’’
What impact will this project have on your customers, your company or the industry as a whole?
Industry 4.0 initiatives have brought many ways to collect data from different machines in the factory. However, the data generated is not often understood or used effectively.
With the proper data architecture in place and a strong understanding of what we’re collecting, we can make better data analytics models to plan and predict factory schedule and capacity. We can also build both internal and external-facing tools to improve the efficiency of the manufacturing process.
We are helping our teams more easily identify quality issues as early as possible so they can make improvements as needed.
