Laboratory for Analytic Sciences
Tools
Figma
After Effects
Miro
Role
UX/UI Designer
Objective
Design an interface using the affordances of machine learning to provide a personalized user experience so that intelligence analysts can quickly and knowledgeably enter their day’s workflow.
Collaborators
Elizabeth Gabriel — Designer
Jillian Swaim – Designer
Problem
Intelligence analysts have been tasked to provide indications and warnings of potential terrorist actions against US interests. Many of the terrorists have never been identified, and are scattered all around the globe. They are members of different but often linked organizations and can be called upon at any time to unleash acts of terrorism of many different sorts. Chloe is faced with lots of "dots" in the form of intelligence data that come from many different sources. Her first task is to decide which dots she should be trying to connect and then to make inferences about what these connections may mean. This is the basic step in generating hypotheses about what these connected "dots'' might mean.
Process
Before heading into Figma, Miro was our home for collaborating on notes and research
We started by developing a persona for Chloe, the analyst reporter. We were very specific about her work style, personality traits, and motivations. From there we constructed two task flows, making sure to identify potential pain points.
Chloe's persona and task flows
Chloe's task flow
Next we conducted interviews with five real intelligence analysts. I have blurred the faces of the interviewees out of respect for their privacy.
We jotted down some key take aways from these interviews and started to brainstorm some opportunities for improvement.
Takeaways and opportunities
We did a lot of independent sketching before coming back together as a team.
Pulling from that batch of sketches, we started to refine some of them into more complete ideas.
Spin Style Queries concept
Lots of whiteboarding sessions
Prototypes
The final prototypes were created in Figma
We created a variety of high fidelity mockups before settling on a direction.
When the analyst gets to work in the morning and opens up their Atlas software, they see a summary of their day's work.
All new reports are summarized by AI in a running text on the left side of the screen. Pre-highlighted words are paired with images and external documents arranged on the right. Emails, customer interests and tag filters are arranged at the bottom of the screen.
Homepage
When the analyst gets to work in the morning and opens up their Atlas software, they see a summary of their day's work.
All new reports are summarized by AI in a running text on the left side of the screen. Pre-highlighted words are paired with images and external documents arranged on the right. Emails, customer interests and tag filters are arranged at the bottom of the screen.
Homepage
The analyst can scroll the report and highlight new key phrases. As they highlight new phrases, new documents (images, reports, coordinates, videos, etc.) appear to the right.
AI will automatically change out tags that are relevant to the highlighted phrases.
Scrolling through the document
The analyst can enlarge images and see details, such as text, highlighted by the AI.
Enlarged image
Selecting a tag will open ll relevant documents to the right. By default, the most important six documents will appear and other documents will stack farther to the right.
The AI will also pull up the most relevant emails, customer interests, and notes at the bottom of the screen.
Enlarged image