Yara AnomaliSense™

Note: For confidentiality purposes I cannot describe the product in detail or show clear pictures of the work done. I will focus on the high-level design process and main outcomes. 

​

​I have been working with this product since July 2023 implementing multiple small to big, complex- and not so complex features. The one I would like to highlight in this case is grouping of alerts. 

​

Generally, the process of a new request starts with a user story assigned to me by the product manager. This user story contains a description and acceptance criteria. This information provides me a basic understanding of what needs to be delivered but it can evolve and change as I progress with the exploration of the request. 

​

In AnomaliSense there was already a solution in place for addresing alerts,  however users had to address alerts one by one and they were not satisfied about the way this was done.  They found it time consuming. Therefore, the hypothesis we were trying to validate to help users reduce the number of actions and the generation of too many alerts was "if we group the alerts based on certain criteria, can the user address alerts by groups?".

​

The steps I followed to complete this product improvement were the following:

​

1. Mapped  existing user workflows​​ to help visualise what can currently be done by users in AnomaliSense.

2. Prepared and conducted multiple brainstorming and exploration workshops together with developers, data scientist and product manager. Some of the questions raised during the workshop were:

   • Will grouping of alerts affect the already implemented filtering solution​​?

   • How many sub-alerts should be per group?

   • How will the grouping of alerts affect the other pages within AnomaliSense?

   • When does the grouping of alerts start and end? 

   • How will the email report change?

   • Is this technically possible? 

   • Are we bringing a better experience to users by grouping alerts?

3. Once we all got the answers to our questions and there was a good level of clarity and feasibility of the potential solution, I started drafting some low-fidelity wireframes. 

4. The potential ideas were presented to my team for feedback to continue exploring what was possible to achieve. We would normally ask ourselves, can this be done technically? do we have the necessary UI components?. This approach helps up reflect if we need to make any changes or we have any blockers.

5.  Later on I conducted research via a video call and email with users. One of the research methods used was the A/B testing. The purpose was to understand if the information showed in the card group allows the users to understand what they will see within that group of alerts. What is relevant for them to know? 

6. As we received continuous feedback, I iterated 4 times to find the best solution possible for the users. 

​

Below some pictures of the design and creative process.