How the interview engine works

We believe that if you’re going to trust a tool with your inner world, you deserve to understand how it works. This page explains the mechanics behind Fatement, without jargon and without overselling.

The archetype framework

Fatement uses seven archetypes to map the forces driving your current situation:

• Achiever: worth tied to output, performance-driven identity
• Judge: internal critic, comparison, shame loops
• Wanderer: identity in transition, searching, unmoored
• Builder: desire to create, agency, need for tangible progress
• Lover: connection-oriented, relationship as identity anchor
• Hermit: withdrawal, recovery, need for solitude under stress
• Trickster: pattern-breaker, restlessness, creative destruction

These are not personality types. They are not fixed categories. They are contextual forces that show up in your language when you describe what you’re stuck on. Most people have 2-3 archetypes active at any time, often in tension with each other (Achiever pulling toward action, Hermit pulling toward retreat).

This framework is not a validated psychometric instrument. It is a reflective tool inspired by Jungian archetypal patterns, adapted for practical self-examination. See our disclaimer.

The interview process

Each interview is a short, page-based conversation (one question at a time, not a chatbot stream). The process is adaptive: questions change based on what you say. If you’re being vague, the engine pushes for specifics. If a pattern emerges, it follows up.

Safety monitoring runs throughout. If your responses indicate distress beyond what a reflection tool should handle, resources are surfaced.

AI processing

Fatement uses AI models from Anthropic (Claude) to analyze your responses and generate results. The AI provider does not use API inputs to train their models.

When the AI produces low-confidence results or fails, we tell you. Fallback results are clearly marked as approximate.

Artifact generation

Three types of artifacts are generated from your interview:

• What’s Intact: surfaces what’s already working in your life before diving into what’s broken. Grounded in evidence from your own words.
• Pillar Card: a core insight with supporting evidence and reframe lines. The reframes use your language, not generic advice.
• Scoreboard Rewrite: replaces the outcome metrics you’re measuring yourself against (that may be broken or unreachable) with input metrics you actually control.

What we don’t do

• We don’t use your data to train models
• We don’t compare you to other users
• We don’t benchmark you against norms
• We don’t generate diagnostic labels
• We don’t validate or praise your answers

The engine is deliberately anti-validation. It won’t tell you what you want to hear. It will use your own words to show you what you’re actually saying.

Limitations

This approach has real limits:

• The quality of the output depends on the honesty of your input
• A 15-minute conversation captures a snapshot, not a full picture
• The archetype framework is one lens among many
• AI pattern recognition is not clinical insight
• Cultural context and lived experience matter in ways AI cannot fully account for

We name these limits because transparency is part of the design, not an afterthought.

Questions about our methodology

We welcome scrutiny. If you have questions about how the engine works, email hello@fatement.net.