Five experiences. One idea: you are already thinking. The system just needs to listen.
Second Brain is designed around a belief: the problem with personal knowledge management isn't that people don't have good ideas — it is that the tools make organizing ideas the work. Second Brain removes that friction entirely.
Screen 01 · Brain Stream
Capture is the only job
A three-column layout: left navigation by type and source, centre feed of thought entries, right panel for AI-surfaced patterns and active projects. The AI entity chips (Idea / Goal / Decision / Memory / Task) appear automatically — the user never chooses. The AI reasoning row shows its work inline, building trust rather than hiding it. The compose bar is the only action in the interface.
Screen 02 · Thought Universe
Your mind as a landscape, not a list
Five organic clusters on a warm off-white canvas, each representing a major theme cluster. Nodes connect across clusters with dashed lines showing unexpected relationships. The self node sits at the centre, not a sidebar. A detail panel on the right shows AI-generated insight about the selected cluster. This is the screen that makes the product feel alive — a visualisation of someone's actual thinking patterns.
Screen 03 · Decision Studio
Your own thoughts as evidence
Three columns: evidence gathered from your Brain Stream (left), a collaborative thinking canvas in the centre (not a chatbot — a conversation with your own data), and an emerging conclusion panel on the right. The AI never advises — it surfaces what you already know and asks better questions. The weight bars visualise evidence balance visually, making the cognitive work legible without numbers dominating.
Screen 04 · Memory Replay
Not search — narrative reconstruction
A natural language query bar, an introductory narrative paragraph (the AI synthesises context), then a vertical timeline with event cards from multiple sources — typed thoughts, voice notes, meetings, decisions. An italic section header separates chapters. Each card shows the source type, date, and an AI annotation connecting events. The sidebar shows suggested queries, making the feature discoverable without being a search engine.
Screen 05 · Future Paths
Futures you can modify
Four paths in the left panel, main view shows the selected path with a trajectory chart (three lines: current, with-AI, plateau risk), six dimension cards with projected changes, a milestone timeline, and a modify-assumptions section. The assumptions section is the key interaction — toggle each assumption on/off and the trajectory chart updates. This makes the product feel responsive to agency, not fatalistic.
System · Visual Language
Calm intelligence over visual drama
Instrument Serif (display) and Geist (UI) create an editorial-meets-digital register — thoughtful, not techy. Warm off-white base (#faf8f4) reads as premium and personal. Amber (#c8840a) is the single primary accent — used sparingly for AI presence and active states only. Six semantic colours for entity types are consistent across all five screens, building an intuitive visual vocabulary across the product.