Ayasdi Iris

Iris’s mission is to democratize machine learning and make enterprise level artificial intelligence available to all.

Make AI approachable through simple visualizations.

Ayasdi’s current platform requires advanced knowledge of python, has a complex interface, and a steep learning curve. Iris is designed to let people with little data science background simply upload data and quickly gain insights into that data.

Automate complex tasks

Using a series of questions and steps, Iris guides the user through  encoding data set creation. Automated charts are linked with AI visualizations to facilitate rapid discovery and justification.

Create Intelligent Apps with ease and simplicity

Iris created apps let data reveal its natural structure and relationships. UI features facilitate a rapid understanding of segments, anomalies + hotspots.

We began sketching out the journey, brainstorming and getting ideas onto paper quickly.

Since we had some desired outcomes and a good idea of the user journey from the existing platform, we began to ideate how the Iris journey would differ and began to map the paths by creating User Flows. These became a map of the user journey where we tried to identify pain points early on and to eliminate them. This would also become the basis of the wireframes that would scaffold an early click-through prototype that we would immediately begin testing.

Initial PM ideation Balsamic wireframes based on emerging user journeys.

Early wireframes for click-through flow testing.

Partial view of wireframe with sketch prototype click paths shown. At this point, the testing prototype is only fourteen screens with three defined user paths yet is already giving us valuable feedback from testers.

For the sake of brevity, we’re going to jump ahead a month or so. One of the great things about working with Sketch is how easy it is to begin working on visual design and user interface while still refining user experience and flow. I’ve added much more depth and detail to the initial wireframe scaffolding. The screens are starting to look like a fully fleshed product while also still being available for testing and presentation, both internally with Ayasdi’s data scientists but for current and prospective customers as well.

At this point, the prototype is 58 screens with a clearly defined user journey for a basic user as well as entry and exit points for advanced users. A dashboard and project card UI has been added to the experience as well as planned interactions with topological maps and charting.

The latest iteration of Iris with a detail of Ayasdi topological data model. The visual design language is resolved and complete. This detail also shows the non-linear task navigation in the four cards across the top of the screen. Depending on the project the user is building the number of tasks and corresponding tasks can vary. Iris builds all the tasks in parallel; the user can check on progress and even begin work on completed operations while others are still rendering.

An important note; color is used very strategically in both AML and Iris. Orange is used judiciously and only for calls to actions, warnings, and important status updates. Bright cyan is for activity and secondary functions.