← BackAI Product Design · 2025

ReflectLoop

Turn your journal into a mirror

Product Design · Full-Stack · 2025

Open the live prototype
ReflectLoop

What it is

A journaling product where AI reads between the lines as you write, surfacing cognitive biases, recurring beliefs, and thinking patterns directly inside your own text. Voice, photos, video, and a longitudinal pattern timeline build a portrait of how your mind moves over weeks and months.

Problem

Journaling apps store text. They don't read it.

Most journaling tools are blank pages with a save button. ReflectLoop reads between the lines, surfacing cognitive biases, thinking patterns, and recurring beliefs as you write.

AI reads as you write

AI reads as you write

Cognitive distortions are highlighted inline, overgeneralization, should statements, catastrophizing, with gentle labels, not clinical jargon.

Patterns emerge over time

Patterns emerge over time

Weekly and monthly views surface recurring themes. Are you always catastrophizing on Mondays? The data knows before you do.

Product

From writing to insight

Journal

Journal

Insights

Insights

Features

Features

Journal view, desktop. Five categories of patterns underlined inline as marginalia, sidebar groups today's detected patterns and links each to its history.

Journal view, desktop. Five categories of patterns underlined inline as marginalia, sidebar groups today's detected patterns and links each to its history.

Insights, pattern evolution stacked over a year, with a written-headline callout: 'Growth Mindset increased 230% while Overgeneralization dropped 87% over 12 months.'

Insights, pattern evolution stacked over a year, with a written-headline callout: 'Growth Mindset increased 230% while Overgeneralization dropped 87% over 12 months.'

Stack & Tools

AI

Claude (Sonnet + Opus)Claude Code

design

Figma

code

Next.js 16TypeScriptTailwind CSSLucidedate-fns

infra

SupabaseAnthropic SDKVercel

Open question

Does naming the bias inside someone's own sentence land differently from explaining the bias in the abstract?

What the data shows

Whether people keep writing after an insight appears, and whether they open the insight panel at all.

What it does not

Nothing here shows whether the writing changed how anyone thought afterwards.

More client and product design case studies at hannadevries.com