“Tana is one of the few knowledge management tools that moves beyond mimicking static text on paper. It takes the computational medium seriously. It gives regular folks access to a set of powerful primitives that previously only developers could touch.”Maggie Appleton
Product designer
Knowledge graph. Write information, not documents.
Tana is a knowledge graph. Just like how a map shows how different cities and towns are connected by roads, a knowledge graph shows how different concepts and entities are connected to each other. This way of connecting information aligns more closely with how the human brain works compared to traditional databases, giving you and your AI a set of capabilities in Tana that are impossible to achieve with other tools.
What is Tana's knowledge graph?
A knowledge graph is a web of things and how they connect; people, ideas, projects, decisions are all linked together instead of locked in separate files.
In Tana, every note, task, and entity you create becomes part of this graph. You write in a simple outline (fast, familiar, no friction), and the knowledge graph quietly keeps all the connections underneath, ready for you and your AI to actually use.
“This is proposing a new fundamental model for computing, a new mental model.”Alexander Obenauer
Explorer of the future of personal computing
Old habits die hard, but the document has to go.
The document format has survived from scrolls, via the printing press and typewriter, to various incarnations of live web docs — all without fundamentally changing much. The reason is simple: documents are straight-forward, familiar and it's just what we've always done.Documents are painfully limiting, and invariably lead to scattered and outdated information, time wasted copying and pasting across different tools. They also don't lend themselves well to leveraging the power of AI, because of their lack of proper structure.Why is it that we have to choose between the familiar flexibility of a plain document, and the awkward rigidity of structured tools? At Tana, we believe we have found a sweet spot.
Knowledge graph + outline editor = ❤️
An outline editor is the perfect interface for working with a knowledge graph because it allows for efficient navigation and organization of complex information. However, it has lacked a way to represent structure in your information, beyond a simple hierarchy. Tana has brought two important innovations to the outline editor that solve this issue: supertags and their fields let you capture structured information, and views let you visualise information within the outline editor.
“Tana's design reflects deep knowledge of the strengths and weaknesses of existing connected outliner tools.”Maggie Delano
Assistant Professor of Engineering
“Tana operates in the sweet spot of outliners and databases, creating a new kind of flexibility that sometimes resembles magic.”Alexander Rink
Computer scientist
Knowledge graphs are superior for AI.
As we are entering the AI era, a lot of things will change. The single-purpose software that we've been used to, with their rigid interfaces and traditional databases, will struggle to leverage the new possibilities.Knowledge graphs are better suited for AI than traditional databases because they provide a more versatile and expressive way to represent and connect data.Graph-based data models allows for more complex relationships between your information, which in turn helps AI systems to better understand and reason about it. They can be updated and expanded with ease, making them perfect for ever-changing information and needs.Learn more about AI in Tana
“Tana is to the knowledge graph, what Netscape was to the Internet.”Torbjorn Nerbovik
Tech Lead @ Tikkio
What you can build on Tana's knowledge graph
Because everything lives in one connected graph, you can build lots of different "apps" on top of the same information:
The nice part? You don't have to decide up front whether something is a "note" or a "task" or a "record." It's just a node in your graph. You can look at it however you need to, whether the use case is reading & research, thinking & writing, tasks & projects, life OS / second brain, or something else.
Questions and answers
Can I get my data in and out?
Yes. Currently we support Roam, OPML/Workflowy and Logseq as native import formats. Export as Markdown or JSON, or check out Tana Publish, whenever you need to move things elsewhere. Your data stays yours.
Why do knowledge graphs work better with AI?
Documents are flat, AI sees a wall of text and has to guess at relationships. Knowledge graphs are structured, AI can follow explicit connections, do multi-hop reasoning ("find all tasks for projects owned by this person"), and ground its answers in real relationships. Less hallucination, more accuracy.
Is Tana a graph database?
Under the hood, yes. Tana uses a graph-structured model. What you see however is a thinking and productivity tool, not a raw database. You don't write traditional queries; you use search, views, and AI to work with your graph naturally.
What's the difference between a knowledge graph and linked notes?
Linked notes (like [[backlinks]] in Roam or Obsidian) show that two pages reference each other. That's useful, but limited. A knowledge graph adds types and relationships: This node is a #person, that node is a #project, and the connection between them means "responsible for." Types let you query, filter, and reason over your knowledge, not just click around.
How is a knowledge graph different from folders or a database?
Folders organize files in a tree; one place per thing. Databases organize rows in tables; rigid and separate. A knowledge graph organizes things and their connections, so the same piece of information can link to many others without duplication. It's closer to how your brain actually works.
Do I need to understand graph theory to use Tana?
Nope. You work in a familiar outline: Bullets, indentation, keyboard shortcuts. Tana handles the graph behind the scenes, turning your writing into a connected structure you (and AI) can search, query, and reuse. No query languages required.