Deep dive
Zeno's brain is a local-first AI second brain — a personal knowledge graph with episodic memory and semantic search. It captures what you do, connects meetings to code to notes, and answers in context instead of treating every chat as day one.
Traditional second brain apps force you to file things. Zeno captures passively. You talk, work, browse, copy, meet — and the brain connects the dots. A photo from Tuesday links to the meeting note from Wednesday links to the task you created on Friday.
This is episodic memory: not just what you saved, but what happened, when, and why it mattered — an AI that remembers everything you opt into, under your control.
Under the hood, everything becomes a node. Conversations, tasks, notes, PDFs, calendar events, clipboard entries, code files, browser history, photos — each a vertex in your personal knowledge graph AI layer.
Edges form automatically: "this note mentions that project", "this task came from that email", "this code change relates to that conversation." Over time, the graph becomes a map of your digital existence — connect meetings, notes, and tasks without manual linking.
Every memory object gets embedded into a local vector index. Semantic search on personal notes means you don't need exact filenames or tags. Search by meaning: "that article about Rust async patterns I read last month" or "the meeting where we discussed pricing."
Embeddings run locally — RAG for personal documents on your machine. Your semantic index never leaves your device unless you opt into cloud sync.
The brain maintains a continuous timeline of your day. Morning brief surfaces priorities. As you work, events stream in — meetings, code sessions, clipboard captures, browser tabs, documents opened. By evening, Zeno has a complete picture.
This timeline isn't a log file. It's structured, queryable, and connected to the knowledge graph. "Show me everything related to Project Atlas this week" returns a narrative, not a dump.
Screenshots, photos, and visual captures become first-class memory objects. Linked to the projects, conversations, and tasks that were active when they were taken. Vision models (local or cloud) extract context without you writing a caption.
A project in Zeno isn't a folder. It's a cluster in the graph — tasks, notes, code, meetings, documents, and conversations that share context. Switch projects and your companion's awareness shifts. Ask about one project and Zeno doesn't confuse it with another — an AI assistant that remembers projects distinctly.
The brain runs on a local SQLite spine with optional cloud sync. Three tiers of AI: bundled local model for offline use, Ollama for primary local inference, bring-your-own-key cloud when you need more power. Memory, embeddings, and search default to local. Cloud is opt-in, not opt-out.
When you're away, the brain doesn't sleep. Background cycles consolidate memories, detect patterns, surface forgotten connections, and prepare tomorrow's brief. Habits emerge from observation, not manual tracking. "You usually code best between 9–11am" isn't a setting — it's learned.
The long-term vision: Zeno becomes the narrator of your digital life. Not a diary you write, but a story that writes itself from everything you do. Years of projects, learning, relationships, and growth — queryable, visualizable, and always under your control.
Software that passively captures your digital work, structures it in a graph, and lets you ask natural questions across notes, code, and meetings — locally when you want privacy.
Obsidian and Notion are note-first. Zeno is companion-first: tray presence, voice, automation, and memory that links sources you didn't manually file. Compare tools.
Yes. Ingest PDFs and documents into the graph, embed locally, and query with Ollama or other local models — no cloud required.