Storytrack — Designing a decision clarity layer for product teams

In most organisations, decisions don’t vanish — they fragment.

A research insight lives in a slide deck. A trade-off is buried in Slack. A stakeholder conversation gets summarised differently by three people. Weeks later, when someone asks “why did we ship it this way?”, the team either:

• re-litigates the decision (wasting time and reintroducing bias), or
• reconstructs a plausible story from memory (which quietly becomes “truth”).

Most tools are optimised for writing (blank pages, rich formatting, infinite flexibility). But the real pain shows up later — when a new joiner, stakeholder, or even the original team needs to reconstruct causality: what we learned → what we decided → what changed → what happened.

Storytrack is my attempt to solve that specific gap: a lightweight way to capture turning points as they happen, and retrieve them later as a coherent narrative — not a folder of artefacts.

I built Storytrack as a solo founder, wearing the Senior Product Designer hat end-to-end: I defined the product model, information architecture, flows, interaction patterns, and microcopy — and shipped a working MVP.

AI-Assisted Development

I don’t write production code manually. I used AI-assisted development deliberately: I wrote detailed specs and behaviours, defined the data model to support the UX, had the AI implement, and then I reviewed and iterated through real usage and QA. I treated the AI like a development partner: I directed the work, I controlled quality, and I owned the product decisions.

Stories and Tracks

A Story is a project/initiative narrative over time. A Track is a turning point captured at a moment in time.

I constrained Tracks to four types:

This is the core product decision: structure that’s strict enough to be scannable, but light enough to not feel like process theatre.

Timeline-first

The story page is designed for two modes: Authoring (minimal friction) and Reading (scan narrative, filter by type). A document-first tool asks you to write a summary of reality. A timeline-first tool makes sequence and causality visible by default.

Optional metadata

Each track supports tags, links, and attachments — but none are required. The MVP track is just type + title + a few lines. Metadata is there for credibility when needed, not as a mandatory ritual.

Search as a "trust feature"

If someone asks "didn't we already decide this?", the product needs to answer immediately. I treated global search (Cmd+K) as core: accessible from anywhere, searching across stories/tracks/tags.

Sharing that respects reality

Stakeholders often won’t have accounts. Storytrack supports a simple model: generate a tokenised public link (read-only) that can be revoked. This satisfies the "stakeholder needs access" and "team needs control" constraints simultaneously.

AI summary

I positioned AI as assistive, not central. It generates a concise narrative recap from existing tracks, helping new joiners get up to speed without reading every entry. It's gated (only on populated stories) and persisted.

Personas

Flows and Diagrams

Early Exploration

Live App