Elise Marchetti, Editor-in-Chief, AI Companion Testing

Elise Marchetti

Editor-in-Chief, AI Companion Testing

Lisbon, Portugal · she/her · Covering this beat since 2022

Has been testing AI companion apps in long, unglamorous stretches since 2022, and sets the scoring rubric every review on this site runs through.

2022

Testing since

40+

Apps scored

600+

Hours logged

About

Elise Marchetti runs the editorial side of My Intimate AI. She started testing companion chatbots in 2022, back when the category was mostly two apps and a subreddit, and she has kept the same habit since: set up a character, talk to it every day for three weeks, and write down the exact message where it stopped feeling like a person. The first night is easy to fake. The twentieth is not.

Her background is in consumer software reviewing, which is where the rubric came from. Every app that gets covered here is scored on seven axes — opening quality, memory retention, persona drift, refusal behaviour, pacing under a slow-burn prompt, tone control, and what happens when you contradict it. The same five scripted scenarios go into every app, so two reviews written eight months apart are still comparable.

She is also the person who decides when an app does not get covered at all. If a service cannot state who processes its messages, or its cancellation flow is a support ticket instead of a button, that goes in the review body rather than getting quietly dropped. She has logged more than 600 hours of companion chat across 40-plus services, and she still keeps the transcripts.

Areas of focus

  • Conversation quality and persona consistency
  • Long-term memory testing across multi-week sessions
  • Romance pacing and slow-burn roleplay
  • Review methodology and scoring rubrics
  • Character creation systems

Background

  • Testing AI companions since 2022
  • 600+ logged hours of companion chat
  • 40+ services scored on the same seven-axis rubric
  • Sets the editorial standard for every review on this site

Articles by Elise Marchetti

Fact-checked by Elise Marchetti

The whole cluster hangs off how to judge an AI companion site, which is the piece this team points people at first. The companion chat this team runs is the site the blog belongs to.