The way CEO Tyler Jewell tells it, customers had to bang loudly on the door to get any AI action out of Akka – a widely deployed middleware toolkit for building concurrent, distributed, resilient Java applications.
“We had to set aside our biases” he tells The Stack – saying that the company had been sceptical of the endless AI and “agentic” market hype.
But now it’s betting the farm on agentic AI, after a steady drumbeat of customers saying they were already using Akka to build agentic systems and needed more capabilities forced it to extend what it could do, he says.
Akka this month released its Agentic Platform, featuring a “single SDK and runtime with all the components necessary to create autonomous AI systems: orchestration, agents, memory and streaming” as the firm has it.
That platform, crudely, is a Langchain alternative for Java applications; a Swiss Army knife for supporting high-volume agentic AI workflow automation, which Jewell says lowers compute costs dramatically compared other approaches to building agentic applications.
