Year
2022
T-Hex: The Analytical Task Assistant
Collaborators
Katie Denson
Jacob Williams
This video mockup was completed for LAS as part of an initiative to bring AI into data analytics work. We conducted months of interviews, attended AI workshops and lectures, and ultimately created demo videos meant to guide the AI architects who would begin building software inspired by our investigations.
My role in this work is present throughout, but I’ll start by saying I fully animated this and all other mockup demos myself. I’m most proud of my work as a team leader — I worked to set deadlines and communication standards that held myself and the rest of the team accountable, and the work was better for it. Of course I’m also proud of the expansive ideation process (my favorite part of any project) but that work is harder to divide into individual moments, especially because it’s a space where you leave your ego at the door.
The narrative in this video requires a lot of context, but in short it is for a language data security analyst working at an unnamed government intelligence organization. Our user is outside of his area of expertise, and trying to get his bearings as fast as possible.
Pain points in current language analysts' task flows included an unclear understanding on how to approach such massive datasets, an inability to quickly communicate with subject matter experts, and difficulty balancing company interests against the changing landscape of customer interests and context for datasets.
From speculative design
to functional framework:
Below are slides from a presentation about this project I conducted for a series of undergraduate studios. My team spent months conducting interviews with current and former language data analysts; a complicated feat given that most of their knowledge is confidential. With what they taught us we began working on speculative designs, starting with the fantastical and circling inward toward an interface designed to aide analysts gain situational awareness and data at a glance.
An Intelligent, Modular Task Flow
By weighing size, color, and proximity, users can understand both to team priority of tasks as well as the relevance of one task to the next. Tasks can be rearranged at will to show different priority systems, such as "tasks I can get done before my 2pm meeting," or "tasks related to nuclear weapons development."
The system is there to guide the user without being prescriptive. As shown below, there are a number of ways the user might approach the task flow presented here.
Adaptive Information,
Only when you need it
With thousands of hours of untranslated audio, our user is easily overwhelmed. But with adaptive news feeds, information cards, and team updates, the only information they see is the information they want.