Skip to content
REALI

REALI · PR & COMMUNICATIONS

Amplify PR with structured AI.

REALI is an AI work platform for PR and brand teams. Our core product, GEO Positioning System (GPS), helps you understand how AI describes brands and issues.

Use the system, read a report or commission a project to bring those findings into your communications decisions.

AI executionHuman judgment

Build intelligence with every project.

Meet REALI and the team →

Our current core product · GPS

How does AI describe your brand? Start with the answers.

GEO Positioning System uses repeated tests across models to organize answers, visible citations and recurring narratives—giving your team material to investigate.

Recorded operation · Apple example · Not a customer outcome

Read the answer. Inspect the evidence.

Compare model answers, trace recurring narratives and visible citations, then apply professional judgment.

Ways to work with REALI

Use the system, read the research or commission a project.

Three ways to access the same REALI and GPS research foundation. Start with what your team needs today.

  1. Ongoing use

    Annual system access

    For teams with ongoing research needs. Initial purchase includes GPS training and a workshop on reading GEO data.

    Explore annual access →
  2. Public research available

    Research reports

    Explore public studies of how AI answers change. Research membership and paid access to selected major reports are planned.

    Browse research reports →
  3. Scope agreed per project

    Project services

    Bring a question about your brand or an issue. Discuss research carried out with GPS and GEO consulting, with the scope agreed in advance.

    Explore project services →

Featured research

One debate. Two AI models. Do they cite the same sources?

Our Taipei mayoral debate study compares AI answers at three points in time, looking at cited sources, candidate comparisons and changing topics.

It examines AI answers under specific test conditions. It is not an opinion poll or a rating of the candidates’ abilities.

Published · 2026.11.01

From U.S. Elections to Taipei’s Debate: What Shapes AI’s View of the Candidates?

Study period
2026.10.30–11.01
Models
ChatGPT · Gemini

By Raymond Wu and the REALI research team. Findings describe AI answers under the study conditions, not public opinion or candidates’ abilities.

Read the study and charts →

Latest news

Follow what REALI is publishing and developing.

Research publication

How AI answers changed around Taipei’s mayoral debate

Read the full study →

Media coverage & perspectives

Research that opens a wider conversation.

Explore Mirror Media’s reporting and related articles on ETtoday Forum and SETN, alongside REALI’s original study.

Your next step

What does your team need to understand?

You do not need a finished question set. Tell us what you want to learn and which decision the research should support. We can then discuss a suitable approach.

Scroll horizontally to inspect the original. Press Esc to close.