A personal AI agent may store some private data if it offers memory, account history, personalization, or connected features. The exact data stored depends on the app's design and privacy policy.
Stored data might include conversations, preferences, saved memories, created tools, account details, usage patterns, or files you upload. Some apps store more, some store less, and some offer controls to delete or limit memory.
You should check the official privacy policy and settings rather than assuming. Look for options to view memory, delete history, disable personalization, export data, or close your account.
You can usually confirm storage behavior directly in the product. Open settings and look for a memory or personalization page; a trustworthy agent shows what it holds and lets you remove items line by line.
Distinguish the layers of storage: chat history, extracted memories, and account data are usually separate systems. An app may let you delete one while quietly retaining the others, so check each individually.
Storage is acceptable when it is purposeful. Let the agent keep the operational details that make it useful, question anything stored without a clear function, and remove entries that no longer serve you.
Kaijie Chen is an entrepreneur and technologist whose expertise spans artificial intelligence, human‑computer interaction and the development of gamified social products. He combines a rigorous academic foundation with wide‑ranging technical skills and has repeatedly demonstrated the ability to translate research into successful commercial ventures.
In 2013 Mr Chen earned the Gold Medal in Physics at the global Yau Science Awards, an honour that underscored his exceptional command of both physics and mathematics and brought him national recognition. Beginning in 2015 he pursued mechanical engineering at Duke University, specialising in human‑computer interaction. During this period he joined Professor Mary Cummings’ NASA Mars‑rover interface project, where he analysed how interface design affects operator trust, confidence and decision‑making; the work was commended by the university. After a period of leave to focus on industry experience, he returned to Duke and graduated in 2020 with a near‑perfect GPA of 3.99.
Between 2017 and 2018 Mr Chen interned at Zhihu under former Chief Technology Officer Li Shenshen, leading data‑driven business analysis and strategic planning initiatives that deepened his market insight and operational acumen. In 2018 he co‑founded an artificial‑intelligence smart‑home start‑up, gaining hands‑on experience in product development, team leadership and market entry while also volunteering for China’s national “Space C” outreach programme.
While completing his degree, Mr Chen served as a product advisor to Xiamen Black Mirror Technology, where he defined the MetaMaker product line. From 2021 to 2023 he joined with Junhong Chen to establish Yunzhongzi Technology, a company dedicated to integrating AI with gamified social interaction. There he created the GPT‑2‑based “Stanford Town” gameplay model, which attracted industry attention and investment from Gaorong Capital. During the same period he was invited to join Duke University’s advisory board for its game‑design programme, providing strategic counsel on curriculum development and talent cultivation.
Since 2023 Mr Chen has co‑founded MidReal, an AI‑powered storytelling platform launched in November 2023 with Boan Chen and Junhong Chen. MidReal has secured investment from MiraclePlus, Linear Capital, Yuan One Capital and ZhenFund, and has already surpassed two million users, demonstrating strong product–market fit and sustained innovation. Now, the original team of MidReal is focusing on Macaron AI, world's first Personal AI Agent. Mr Chen remains committed to advancing the intersection of artificial intelligence, interactive entertainment and user engagement.
