Estimated reading time: 3 minutes
Alan Sternberg, General Manager of Artificial Intelligence at Comply365 recently shared with Aircraft IT an introduction to CoAnalyst and reflected on the original vision and mission that inspired its creation.
CoAnalyst was born in 2023 when we sat down with some of the best aviation safety teams in the world and asked them a straightforward question: what is actually breaking in your day-to-day work? The answers were consistent and, honestly, they were sobering.
Every aviation operation generates enormous volumes of reports. Safety reports, injury reports, damage reports, fatigue reports, air safety reports, etc. filed every day by pilots, cabin crew, ground personnel, maintenance engineers, and staff across the entire operation. And virtually all of that data was being processed manually. Human analysts, working through stacks of reports, applying their own judgment, their own categorization logic, their own sense of what mattered. The result was exactly what you would expect: inconsistent analysis, poor data quality, and a chronic inability to keep up with the sheer volume of information coming in.
But the volume problem was only part of it. The deeper issue was that all of this data was sitting in silos. Safety reporting data over here, FOQA and flight data monitoring over there, technical logs somewhere else entirely. Nobody had a joined-up picture. Root cause analysis at scale was virtually impossible. Trend detection was reactive at best. And there was essentially zero collaboration between safety teams who were, in theory, working towards the same goal. What struck us most was not just that these problems existed, but that they had existed for a long time, and the industry had largely accepted them as the cost of doing business. We did not accept that.
CoAnalyst was our answer. We built it to do what human analysts simply cannot do at scale: apply AI, combining large language models, natural language processing, and machine learning classification, to process safety data consistently, continuously, and comprehensively. The first version reduced the time to process reports by 90%. More importantly, it gave safety teams something they had never had before: the ability to do real-time event analysis and genuine proactive trend detection across their safety data. Not just looking at what happened last month but actually seeing what is building right now. The vision from day one has been simple. AI sees everything in the data. We want to make sure organizations can too.
“AI sees everything in the data. Our job is to make sure organizations can see it too.” – Alan Sternberg
To continue reading this article from Alan, view the original publication here.
