The Coworker Dictionary

A list of actions that an Agent has identified it needs to take in order to complete its task. As the Agent works through a job, it plans out each step — searching a tool, reading a document, sending a message, etc. — and queues them up in sequence. You can often see this queue in real time as the Agent is running, giving you visibility into exactly what it's doing and why.

An AI assistant you build once and let run on its own. Agents can be scheduled to run automatically (e.g., every morning at 9am), triggered by an event, or run manually whenever you need them. They can pull data from your connected tools, generate reports, send Slack messages, and much more. Think of an agent as a tireless team member that handles a recurring task for you — without you having to ask every time.

The no-code interface inside Coworker where you create and configure Agents. You describe what you want the agent to do in plain language, choose what tools it can access, and set when it should run. No coding required.

The default chat mode in Coworker. When you ask a question, Auto Mode figures out the best way to answer it — sometimes that's a quick one-liner, other times it does deep research across multiple tools and data sources before responding. You don't need to choose; Coworker picks the right approach for you.

The integrations that link Coworker to your existing tools — things like Slack, Gmail, Google Drive, GitHub, HubSpot, Jira, Zoom, and more. Once connected, Coworker can read information from these tools to answer questions and complete tasks. There are two types:

An add-on feature that gives customer-facing teams (Sales, CS, Account Management) a live, AI-enriched view of every account. It pulls together data from your CRM, emails, meetings, and other connected tools into one place so you can understand what's happening with each customer without digging across five different apps.

Customer Intelligence lives in the Customers section of Coworker's left-hand navigation.

Coworker's most powerful chat mode. When you have a complex task — like writing a research brief, analyzing customer feedback across many sources, or drafting a detailed report — Deep Work kicks in to run a multi-step process, checking multiple tools and data sources before giving you a thorough response. Auto Mode will switch into Deep Work automatically when needed, or you can select it manually.

A log of every time an Agent has run, including when it ran, what it did, and whether it completed successfully. Execution History is your go-to place to check on an agent — useful for confirming it ran on schedule, troubleshooting unexpected behavior, or reviewing past outputs.

The core view inside Customer Intelligence. It's a spreadsheet-style table where each row is a customer account. Columns can show live CRM data, AI-generated insights (like "What is this customer's biggest pain point?"), or notes you add manually. The AI columns are answered automatically by Coworker using everything it knows about that account.

Coworker's set of features for capturing, transcribing, and summarizing meetings. Once set up, Coworker can join your calls and produce a transcript and AI summary shortly after the meeting ends.

Your company's Coworker environment — essentially your organization's workspace. All your connected tools, team members, and AI context live within your Network. Your Network Admin manages the overall setup and determines which data sources and features are available to everyone.

The person (or people) at your company responsible for setting up and managing your Coworker environment. They connect organization-wide tools, manage user access, and configure settings that affect the whole team.

A Coworker bot that joins your meetings (on Google Meet or Zoom) to record, transcribe, and summarize them. It appears in your meeting as a participant called "Coworker Notetaker." After the meeting ends, you'll receive a transcript and AI-generated summary. You can invite it to individual meetings or set it to join automatically.

The technology that powers Coworker's understanding of your company. OM1 continuously synthesizes context from all your connected tools — meetings, documents, Slack conversations, code, emails, customer data, and more — so Coworker can give answers that are relevant to your organization, not just generic AI responses. You don't interact with OM1 directly; it works in the background to make every Coworker response more accurate and personalized.

Preferences and context saved so Coworker remembers them across every conversation. Coworker "learns" these memories as you use the product, but you can also add memories directly if there are certain things you want Coworker to know, such as your communication style or your preferred tools.

To add a memory, just say "remember that..." during any conversation, or go to your profile settings and select Personal Memories.

The message or question you type into Coworker to kick off a task or get an answer. The more specific your prompt, the better Coworker's response. Think of it like giving instructions to a very capable colleague — clear context leads to better results.

A feature in Coworker that lets you find documents, files, and information across all your connected tools at once. Instead of hunting through Google Drive, Slack, Notion, and GitHub separately, you search in one place and Coworker surfaces the most relevant results from everywhere. You can filter by owner, date, or data source.

The condition or event that causes an Agent to start running. Triggers can be time-based (e.g., "run every Monday at 8am"), manual (you click a button to run it), or event-based (e.g., a new message arrives, an API call is received, or an action in an external tool fires it off). Setting the right trigger is what turns a one-time task into a fully automated workflow.

The main screen where you interact with Coworker — think of it as your AI command center. This is where you type prompts, start Deep Work sessions, and access your connected data sources. You can get there any time at app.coworker.ai.