Job hunting is broken. Between endlessly scrolling through job boards, manually tweaking cover letters, and hunting down recruiter emails, the process is exhausting. I realized I was spending more time formatting documents than actually building software.
I decided to fix it by treating myself as the "CEO" and building a digital workforce to do the heavy lifting. The result is Taskmaster AI, an autonomous, multi-agent orchestrator.
(Note: I wrote this article as part of my submission for the #AllThingsAgenticHackathon).
The Architecture: A "Blackboard" Memory System
Instead of writing a single, massive script that breaks easily, I built four distinct AI agents. To make them work together seamlessly, I used a Stateful Memory Pattern. Google Cloud Storage (GCS) acts as the central "brain." One agent writes its findings to the bucket, and the next agent reads from it.
