Tokenomics meets topology - inside Dynatrace's AI-era observability rethink
For most of the past decade, the working assumption in enterprise observability was simple: when in doubt, collect more. Storage was cheap, pipelines were tolerant, and data quality could be sorted downstream. That assumption is cracking under the weight of Artificial Intelligence (AI), and the bills are starting to arrive.
Josh Clay, RVP of Solutions Engineering at Dynatrace, has been in enough customer conversations over the past few months to know what's driving the anxiety. It is not, he says, a failure of ambition. Organizations are experimenting, shipping, and in some cases moving into production faster than they expected. AI economics, however – driven in large part by Large Language Model (LLM) token consumption – are fundamentally different from the economics of traditional software observability, and most enterprises are discovering that after they commit.
Clay says:
There's still a bunch of AI illiteracy out there right now. You see somebody...
Copyright of this story solely belongs to diginomica.com. To see the full text click HERE
