Invoko

Product Introduction

1. Definition: Invoko is a native AI desktop assistant application designed specifically for macOS (Apple Silicon). It operates as a contextual, on-demand layer that integrates directly with the user's active applications and on-screen content.
2. Core Value Proposition: Invoko exists to eliminate context-switching and accelerate workflows by providing an AI helper that can see, hear, and act upon the work you are already doing on your Mac. Its primary function is to allow users to talk to their screen and have tasks executed across applications without leaving their current flow.

Main Features

1. Contextual Voice Invocation ("Hold Fn"): The primary interaction model is a universal, hold-to-talk shortcut (defaulting to the Fn key). When invoked, Invoko uses the macOS microphone permission for that single instance to capture a voice command. It then processes the speech-to-text to perform actions like replying, rewriting, answering, or dispatching a task, all without a persistent listening state.
2. Live Screen Context Analysis: Upon invocation, Invoko can analyze the current application state, including the open app, window title, URL, selected text, focused input field, and a screenshot when necessary. This allows the AI to understand the immediate context and provide relevant responses or take actions on the visible content, such as summarizing a document you are reading or drafting an email based on a conversation thread.
3. Cross-App Task Execution & Handoff: Invoko can perform multi-step tasks that span multiple applications. A user can dictate a complex request, approve Invoko's proposed action plan, and the app's agent will navigate between authorized applications (e.g., email, calendar, Slack, browser, design tools) to gather information or complete actions, delivering the final result back to the user. This turns short voice commands into longer automated workflows.

Problems Solved

1. Pain Point: The constant disruption of workflow and loss of context caused by switching between applications for simple tasks like replying to a message, searching for information, or updating a document. This "context tax" slows down productivity and breaks concentration.
2. Target Audience: Knowledge workers, designers, developers, and professionals who use multiple applications simultaneously on a Mac and spend significant time in Figma, Gmail, Slack, Notion, browsers, and IDEs. It's ideal for power users seeking efficiency gains and reduced digital friction.
3. Use Cases: 1) Quickly responding to a Slack thread without opening the app. 2) Dictating a follow-up email while reviewing a client's proposal in a PDF. 3) Explaining a feature in unfamiliar software by asking Invoko to interpret the current screen. 4) Summarizing a webpage or extracting key points from selected text on screen. 5) Scheduling a reminder contextually related to the document or task you are viewing.

Unique Advantages

1. Differentiation: Unlike traditional chatbots (ChatGPT, Claude) or voice assistants (Siri) that operate in separate windows, Invoko is a non-intrusive overlay that works within your current application context. It replaces the need for a dedicated chat tab and the associated task-switching. Its "memory" feature, which references past interactions, also distinguishes it from stateless AI assistants.
2. Key Innovation: The combination of on-demand, privacy-centric activation (Fn key) with deep, real-time OS-level screen and application context awareness. It does not require constant listening or a separate interface. Its ability to form and execute multi-app action sequences based on a single voice command, while keeping the user's Mac as the primary workspace, is its core technical innovation.

Frequently Asked Questions (FAQ)

1. Is Invoko always listening or recording my screen? No. Invoko is invoked on-demand via a shortcut, tap, or voice request. It only accesses the microphone, screen context, and performs actions after receiving explicit user permission for that specific instance. Voice recordings and screenshots are not stored by default.

2. How does Invoko use my screen data? Invoko uses your screen context (like the active app, window title, selected text, and URL) only to understand the immediate task you are requesting. This data is used to provide accurate answers or execute relevant actions on the content you are working with, without unnecessary data transmission.

3. What apps and websites does Invoko work with? Invoko is designed to work with common macOS productivity surfaces. It supports interactions with email clients, chat applications, web browsers, documents, notes, code editors, calendars, PDFs, design files, and task management tools. It uses accessibility APIs to understand and interact with these interfaces.

4. Can Invoko perform tasks that involve multiple apps? Yes. For longer or more complex requests, Invoko can use a background agent to move between multiple authorized applications, gather information from different sources (like reading a Notion page and checking your calendar), and compile a result or complete an action, such as drafting an email with details pulled from various contexts.

5. What about privacy and data security? Privacy is a core principle. Invoko is built to minimize data transfer; most processing occurs on your Mac. Your voice is not stored on Invoko's servers, and most data never leaves your device. You have control over permissions for microphone, screen access, and accessibility, and can connect accounts (like Gmail or Slack) only when needed for specific features.