MSSM Student Developing Tool to Facilitate Meaningful Workplace Conversations
By Jess Ignasky
Nearly 64% of digital users under 30 have faced some form of online abuse in their life. Additionally, about 35% of young women reported unwanted interactions as their top concern about using the internet. This problem is even more prevalent in the corporate environment.
Pranathi Alla, Master of Science in Software Management ’22 experienced these pain points herself when she created an Instagram account to share her personal photography. “I created [the Instagram page] to showcase my interest. It was great–I learned how to generate engagement and my page was gaining traction. I was proud of my page and wanted to show it off, so I tagged it on my personal account’s bio section.” says Alla.
“A few weeks later, I was bombarded with random messages and comments on my photography account from total strangers objectifying me because [the page] was no longer anonymous. They knew I was a female.”
Inspired to find a solution for mitigating such interactions, Alla developed the idea for Mannerly in her Start-Up Creation class. “The idea initially started as ShepherdMe.ai, a solution to target the online abuse that women are currently exposed to on digital platforms. When women use these platforms and are exposed to uncomfortable interactions, it leads to hesitancy in utilizing accessible online resources to their fullest potential,” Alla explains.
“However, the problem is much more pressing and subtle within interactions with corporate employees and there is no solution to mitigate them currently. So, I pivoted my idea to focus on this area. In a nutshell, I describe my solution as Grammarly for meaningful conversations.”
The project has already achieved impressive milestones. ShepherdMe.ai was selected for NSF I-Corps, a regional program that uses experiential education to help individuals gain insight into entrepreneurship, and has received a dedicated amount of $3,000 for customer discovery.
“In order to ensure that we were going in the right direction with the project, we talked to customers to analyze whether our hypothesis and assumptions held weight,” Alla states. “We were able to conduct 30 customer discoveries that supported our claims and we moved to the next phase. I understood from the NSF experience that targeting B2C first would be challenging given the high entry barrier and competitors. With this in mind, targeting employees in a corporate office in a B2B offering became my new strategy.”
The new strategy required customer validation, with Alla showing potential customers a visual mockup of what potential collaboration would look like.
“We began having conversations with digital agencies and corporate offices like TCS, Byjus, and more, and spoke to firms that provide HR solutions to potentially partner with them."
Alla had initially worked on a prototype for an abuse verification model using Angular, a component-based framework for building scalable web applications, and the BERT model, a transformer-based machine learning technique for natural language processing. However, with the changes to the solution, a modified MVP is now in progress to fit the new project’s needs.
So, what’s next for Mannerly?
“I am planning to have a modified MVP piloted in Carnegie Mellon University as a DEI initiative to understand how the education system is conducive to this solution and gain valuable insights toward this customer segment. I am in conversation with a potential customer to pilot the MVP in a corporate setting as well,” shares Alla.
“Having a sophisticated model and gathering an unbiased dataset will be a crucial next step for the project.”