Tana AI

Tana is an AI-native meeting and work platform that combines video calls, notes, tasks, and documents into a single connected knowledge graph, with AI agents embedded directly in the meeting experience. Instead of using separate tools for video calls, AI notetakers, project management, and company knowledge, Tana centralizes them so that work happens in the meeting itself: agents listen to conversations, capture key decisions, and create structured data such as tasks, follow‑ups, and documents tied back into the workspace. During live meetings, Tana provides a user-friendly video interface where agents can join, transcribe, summarize, and act on the discussion while participants talk. These agents can generate documents, file issues, update existing tasks and projects, and maintain a memory across sessions, so future meetings automatically build on past context instead of starting from scratch. Because everything lands in a knowledge graph, information like docs, projects, tasks, chats, decisions, contacts, and meetings are interlinked, making it easier to navigate, search, and reuse outcomes across teams. Under the hood, Tana orchestrates leading large language models such as Claude, Gemini, and OpenAI models over this structured context graph, choosing appropriate models for different tasks while giving users the benefits of long-term memory and tailored workflows. This architecture allows agents to reason over prior meetings, related documents, and organizational structure, making follow‑up actions more accurate than generic AI notetakers. The product is positioned to replace or augment a company’s meeting platform, wiki, document system, project management tools, and standalone AI assistants by unifying them in one agentic environment. Tana offers a Free plan for basic meeting memory and AI-powered summaries, along with Plus and Pro tiers that expand AI usage and advanced features. Users can start on the Free plan or a free trial, then upgrade to Plus or Pro directly from account settings as their AI meeting workloads grow. This makes it suitable for individual professionals experimenting with AI-assisted meetings, as well as teams that want to operationalize meetings as a primary place where work is captured, coordinated, and executed by agents.

• ✓Running project status meetings where AI agents capture decisions and automatically create and assign follow-up tasks
• ✓Hosting customer discovery or sales calls with automatic transcripts, structured summaries, and CRM-style follow-ups stored in the knowledge graph
• ✓Conducting recurring team rituals such as standups, sprint planning, and retrospectives with agents maintaining history across sessions
• ✓Replacing standalone AI notetaker tools by letting Tana’s agents transcribe and summarize every meeting directly in the same workspace
• ✓Maintaining a connected wiki of docs, tasks, and decisions that is continuously updated from live meetings

Start by creating a Tana account and choosing the Free plan or a trial of a paid tier, then set up your workspace with key projects, teams, and recurring meetings so agents have structure to operate on. Schedule or start meetings from Tana’s interface and use its built-in video calls so AI agents can join, transcribe, and process the conversation in real time. During meetings, explicitly call out decisions, owners, and due dates—agents are more reliable when responsibilities and timelines are stated clearly. After the call, review the generated summaries, documents, and tasks in the knowledge graph, linking them to relevant projects, docs, and contacts so future meetings benefit from this context. As usage grows, monitor AI usage and upgrade from Free to Plus or Pro if you need more AI capacity or advanced agent workflows, managing your plan through the account settings and billing pages.

Tana’s strongest value is in meetings run through its own video interface; external calls that are not hosted in Tana may not benefit from the full agentic workflow without additional setup. Heavy reliance on orchestrated third-party models (Claude, Gemini, OpenAI) means uptime, latency, and capabilities can be affected by those providers, and model behavior can change as they update. The system’s power depends on having a well-structured workspace and consistent tagging of projects, contacts, and decisions; users who do not maintain structure may get less accurate or less useful automation. Free-tier AI usage and possibly some features are limited compared with Plus and Pro, so intensive users may quickly hit caps and need to upgrade. As with any AI meeting tool, transcription and summarization can misinterpret speakers or miss nuanced context, so human review of critical outcomes remains necessary.