Tensorflow
Producer
UX Researcher
Content Strategist
2019 - 2020
TensorFlow is Google's open source machine learning platform and was first released to the developer community in 2015, but without any marketing guidance. Although it quickly grew and became well loved, TensorFlow lacked a consistent product narrative or visual appearance since the product never had design or messaging strategy.
​
After 2 years of organic, unfettered product growth, I led efforts from the marketing side to introduce brand guidelines complete with new logo, design language, brand architecture, and a messaging strategy. I led the design efforts to completely overhaul the website architecture and design in time for the 2.0 product release at the annual developers conference, shaping how developers interact with the product at every touch point.
Web UX/UI Design
Tensorflow.org originally launched without strategic oversight, and the organic growth led to a convoluted, confusing experience for users. Two years later, our challenge was to take the existing 30k pages on the website and reorganize them into a website architecture that made sense. I lead these efforts, including user testing, rapid iteration prototypes, website architecture, design explorations, and development.
Brand Design
TensorFlow was launched as an open source project by Google developers, and after two years of growing in popularity, the marketing team stepped in to help shape the brand and product vision. From scratch, I lead on creative efforts to define the brand architecture, messaging pillars, design language, a new logo, and ultimately a comprehensive set of brand guidelines to structure the brand appearance in any external outputs.
Developers Conference
I led creation of new branding from the client side, working closely with agency partners. When developing the Quantum AI Brand, we wanted it to speak to numerous identities, often at opposition with one another. It had to feel authentic to the team—who insisted we include design inspirations from art installed in the datacenters—while still resonating with a very sophisticated audience of research scientists, who generally prefer dense academic content to highly designed web pages. We pushed through numerous iterations to land on something both scientific and beautiful.
