Many teams are wasting far too much time and energy searching through massive amounts of log data trying to find answers to user latency issues. Metrics data doesn’t help either as it only tells you that there is a problem, not where to fix it. This is why Coveo turned to observability.

Through implementing observability with Honeycomb, Coveo was able to reduce their user latency by 50 percent.

Coveo engineers, Andy Emond and Mathieu Fortier, along with Honeycomb’s Martin Thwaites chat about how they got there, plus topics around:

• Traditional logging methods vs. distributed tracing
• High-cardinality data: analysis and instrumentation
• Reclaiming your development time: spend less resources on researching and debugging