Persistent AI Agents in Academic Research: 114-Day Case Study of Embedded Agent with Durable Memory

arXiv·medium signal

Alzahrani documents a structured 114-day case study (Jan 31 to May 25, 2026) of a persistent AI agent embedded in a real academic research environment with durable memory, local file access, external tools, scheduled routines, delegated roles, and explicit safety protocols. Unlike benchmark evaluations, this examines what happens when an agent operates persistently over months in a real workflow — one of the first longitudinal implementation studies of its kind.