Case StudyIn development2026

Lorebound

A reader-first interactive-fiction platform that pairs curated story worlds with controlled AI turns, persistent memory, and transparent usage.

Role: Product direction, reader experience, AI contracts, and full-stack architecture

TypeScriptReactCloudflare WorkersSupabaseAIStripe
Lorebound project preview

Problem

AI fiction can feel limitless but quickly loses authorship, continuity, and a clear reason for readers to trust the next turn.

Constraint

The platform needs creative freedom without hidden costs, disposable memory, or a generic chat-first reading experience.

Decision

Anchor play in curated story packs, persist memory around reader choices, and use one visible Ink cost table for turns, retries, images, and context.

Outcome

Lorebound keeps the reader inside authored worlds while making AI capability, continuity, and usage understandable.

Story packs

Reader model

Persistent memory

Continuity

One Ink table

Usage

Overview

Lorebound is an in-development interactive-fiction platform with a reader-first point of view. The goal is not unlimited generated text; it is authored worlds that can respond to a reader while remaining coherent, deliberate, and enjoyable to return to.

Product decisions

Story packs provide the setting, tone, and curated material. AI is used where it strengthens the reading experience: evolving a turn, retaining meaningful memory, supporting a custom action, or adding selective media. Ink makes the cost model visible rather than hiding it behind an ambiguous quota.

What I am building

  • A React reader experience backed by a Hono API on Cloudflare Workers.
  • Shared contracts, structured events, and persistence for story state and memory.
  • Story-pack access and a platform-wide usage model designed for reader clarity.
  • A creator-ready foundation that avoids treating every world as a different product.