“We’ve got now advanced to a full end-to-end agent functionality that spans pipeline constructing, information transformation and pipeline troubleshooting,” Yasmeen Ahmad, product supervisor of information and AI at Google Cloud, informed InfoWorld.
Because of this the agent, whereas accepting enter in pure language, can now perceive schemas, study from present metadata, and grasp the relationships between totally different information belongings, enabling information practitioners to interact with it throughout your entire information pipeline lifecycle, she added.
These engagements may embrace asking the agent to carry out duties reminiscent of producing an information pipeline, modifying present pipelines, and even troubleshooting points, since it will possibly analyze code and logs to establish the foundation explanation for an issue and recommend or apply a repair.
