The tech industry continues to struggle when it comes to diversity, equity and inclusion (DE&I). But in an increasingly data-driven era, numbers may play a key role in advancing enterprise accountability, Colleen Tartow, Director of Engineering at data analytics company Starburst Data, told Silverlinings.
Through the entire process of how data is collected, how it functions or how algorithms are developed, “data science has traditionally been rife with bias,” Tartow said. She cited examples like the former AI recruiting tool from Amazon Web Services (AWS) that showed a massive gender bias.
Tartow has long borne witness to the diversity problem in and out of tech, as the only woman graduating in her astrophysics Ph.D. program in the early 2000s. “The women in tech have always been aware of it,” but recent years have “developed a vocabulary” that has established a more widespread awareness throughout the industry, she explained. And now, data may be a crucial catalyst for equally exposing internal diversity shortcomings and driving more inclusive teams to develop less-biased products and customer-facing services forward.
“Data can help us sell this story,” Tartow said. “I think that we're finally at a juncture in technology where we can democratize the data to help tell the story of humans through data. I think it's very important for us to make sure that story is inclusive and make sure that the data itself and the story that we're telling with that data reflects the human story. Without that, what's the point?”
Democratized data for DE&I
According to Tartow, while the industry made some great strides in diversity, the needle has in many ways remained slow to budge. And current macroeconomic conditions create an even more lethargic landscape for diversity, equity and inclusion (DE&I).
For growth to be possible, data needs to be democratized, and Tartow sees how the ever-growing cloud-based structure of data storage helps this process.
“You no longer need an infrastructure engineer to start developing services because [cloud has] enabled the development of all the SaaS platforms and automatically scalable platforms — horizontal [and] vertical,” she explained.
But because data for DE&I necessitates effective and ethical data governance, the cloud presents both benefits and challenges. “Cloud is all about giving more people access to infrastructure, and data lives on that infrastructure. So, sharing data is easier than ever, which means it should be easier to include more people in the data story.”
“It's also a much more complex challenge from a governance perspective because you've got more people accessing it,” she continued. “You've got to keep tighter controls around that data. And so, the governance and compliance of the risk aspects are something you can't really ignore.” This sentiment is one shared by analysts and business leaders — who see a growing need for improved data management.
Still, with the cloud’s ability to democratize data access, Tartow claims it should be used as a KPI or even more preferably as an OKR “trickling down through the organization” to avoid unhelpfully vague DE&I targets that don’t actually instigate meaningful change.
“All these companies say how important DE&I is to them, and you look at their board and there's zero women, zero people of color. [They] put out these pride statements and these [statements for] Black History Month, and it's like, how many Black people are on your exec team?”
Tartow concluded, “The product reflects the developers,” and for everyone to be benefiting from the technologies they make, there need to be diverse perspectives at the core of the work and at the top floor of the business. “You need to have an inclusive workforce building these inclusive products.”
“Data's just ones and zeros at the end of the day,” she said. “We are what actually gives it context.”