Report Maze
The Auditor manager and I immediately knew where to apply AI in this product. All interaction with the application revolves around reports. I mention this to emphasize that this functionality is one of the main features. If you have a question about the data collected by the Auditor, you search for the report that will answer your question. And there are hundreds, if not thousands, of reports. It's like using an encyclopedia: to find an answer, you look at the table of contents, then browse several articles that you think might contain the answer, and continue this until you find what you're looking for—if you don't give up early. As you can see, it's a real headache.
Yes, at some point, the team added a search function. It was a step in the right direction, but as you might know, creating a flexible, understandable, and convenient data filtering system is not that simple.
Known Issues
For example, assistance with creating search queries is one of the most frequent requests received by our tech support. Click statistics showed us that users rarely open specific reports, which is puzzling, to say the least (see the screenshot below). Additionally, from previous usability testing, we know there is a problem with understanding how the filters work.
I had been raising this issue with the Auditor manager for a long time, lamenting that we were almost repeating the same problem in 1Secure, our new cloud solution.
I had the opportunity to study this functionality in detail when I worked on the "facelift" of the Investigation screen in 1Secure, which had the same search issues. During that time, I gathered data on what the functionality entailed, how it worked, and what problems were known. However, the team had already implemented a solution at the architectural level similar to that in Auditor, and they weren't planning any changes in the near future. My proposal to seriously improve the reporting experience was rejected. Understandably so, to make improvements, we needed a deep understanding of the questions users asked of the reports. Since we didn't have that information, it would require conducting research, developing concepts, testing them, and, if successful, likely making significant changes to the product. All of this required time and commitment from people who were not ready for it.
How AI Can Improve Reporting
Users would no longer need to comb through reports searching for answers or struggle with our complex filters, and we wouldn't need to rework them to make them more understandable. Not to mention, it would reduce the load on tech support.
We boldly assumed that AI could significantly impact the learning curve for working with search and reports, especially considering that users spend an average of about 15 minutes per week on our product and often feel lost in it.
Participating in the AI Hackathon
For our solution, we looked beyond a single product. The company had acquired many applications that, despite addressing different IT security tasks, had similar search and reporting functionalities. At least four of them already had such search capabilities, and the fifth team was considering developing it themselves.
Thus, the core of our solution became a reporting platform capable of working with data from various company products.
We even proposed a possible roadmap, which involved gradually connecting more and more products, thereby increasing the value of our platform. With this, we aimed to demonstrate that we could start small and expand over time.
In other words, users no longer need to worry about setting filters manually, as the chat will guide them on where to look next.
Each investigation saves the entire sequence of requested reports, allowing users to trace back through the investigation process to review previous findings or save them for later.
Recognition and Success
The result was both impressive and valuable. We demonstrated how to significantly simplify the lives of our product users with smart reports by enhancing the current solution using AI technologies. Our idea received recognition and won first place at the hackathon.
Bumps in the Road
I don't believe this was the right decision because we have significant issues in this area, and this technology could fundamentally change everything, greatly increasing the value of our products for our many users.
Nevertheless, this did not diminish my faith in AI and its potential to revolutionize human-computer interaction. Therefore, I believe it is crucial to continue seeking solutions that embrace this new paradigm.
Contact me
