Alpha Vertex

A Dashboard Prediction Tool

hp.png
 

Background

Alpha Vertex developed a massive network of information connecting organizations, people, and events across the globe into a single relational map or graph.  From the massive information, PreCog 500 is a powerful machine-learning model incorporating hundreds of data sets to forecast stock returns at multiple intervals using Artificial Intelligence (AI). 

 
 

The Problem

Looking at a massive amount of data is very hard to decipher and analyze for finance industry professionals. Algorithm-based stock advising tools often speak in a complex language that the majority of hedge fund managers can’t decipher, leaving potentially valuable information undiscovered. How might we create a quick and efficient tool that makes the otherwise dense data easily digestible?

 
 

The Opportunity

There is an opportunity for Alpha Vertex's tools and models to extend to its target audience, hedge fund managers. PreCog 500 can be a supplement for their daily intake, along with using the Bloomberg Terminal. 

 

My Role and Timeline

Project Manager and UX Consultant, 2 1/2 weeks


User Research-Interviews

 

We interviewed 3 finance industry professionals to learn more about their needs. We were also interested in their behaviors and interaction with data analysis. 

 
 

KEY TAKEAWAYS

  • None of the users used Artificial Intelligence, but expressed interest in learning how to use it.
  • Users like to look at raw and visualized data, in addition to the relative value data.
  • The data needs to be quick and presentable; color coding is helpful.

Personas

 

From our interviews findings and synthesis, we started to create our persona, Active Amar. His quote is "Don't mess with what works!" in reference to the Bloomberg Terminal. 

 
 

Design Studio

 

We proceed to Design Studio with Mutisya's (Alpha Vertex's CEO) ideas in mind. We started with pen and paper, then moved on to wireframing with Balsamiq. Here we added 10 Sectors from the Stock Market, an Industries section as a marker, Size, Ranks, Tickers, 5D%, Volatility, and Score. Each sector has its own industry which can break down into market cap size and finally the companies' predictions are revealed.

 

 Wireframes and User Testing

 

We designed our medium fidelity wireframes in Sketch and from there we did user testing. The users can choose a sector, an industry, and market size. The data table below would reflect the chosen parameters. Users can also look at ascending or descending order in regards to 5 Day Prediction, Volatility and Score, again based on the chosen parameters. Annotations are also provided.

 
 

KEY TAKEAWAYS from User Testing #1

  • No reset or clear filters button.
  • Find an easier way to access additional pages on the bottom of the page.
  • Make the prediction column more eye-catching.

Iterations

 

We iterated our designs based on user testing findings and did users testing #2. We added more data and time parameters in the table such as cap, date, and end date of prediction. Lastly, we added a page scroll. Annotations are also provided. 

 
 

KEY TAKEAWAYS from User Testing #2

  • Sections (sectors, industries, market cap) aren't clearly defined.
  • Pressed state of buttons are unclear.
  • Need more information or graph or article: "Why should I believe this?"
  • Add info icons and line charts.

Mockups

 

Prototype Video

 

 

Next Steps

  • Add and format a newsletter via email.
  • Add more information (timelines) to the graphs.