DAPPOS launches xBubble AI agent to build and run SOPs

DAPPOS launched xBubble, an AI agent that generates, tests and dispatches task-specific standard operating procedures (SOPs) to convert short user requests into finished images, video, websites and documents. The product is available today with more than 10 core capabilities and two user modes.

xBubble is built around two systems. Bubble Engine automatically creates and evaluates SOPs for defined tasks. It uses AI coding agents to produce multiple solution variants, assemble test harnesses, combine candidate models and tools, and score outputs against task examples and quality criteria. When an approach meets the criteria, Engine promotes it into a reusable SOP.

Bubble Pilot is the runtime dispatcher. It reads a user’s request, determines the task type and routes work to a matching SOP. If no SOP fits, Pilot runs a general-purpose agent. Recurring fallback requests are reported back to Engine for potential SOP development.

xBubble offers two product modes. Bubble Computer is an end-to-end project workspace that opens a sandbox when the system detects multi-step work and loads specialized skills on demand. In a single run, it can research a topic, draft documents, create visual assets, verify claims and produce a final deliverable while managing model selection and step coordination.

Bubble Personal operates across local files, browsers, apps and schedules to automate work that requires personal accounts, produce calendar- or email-based briefings, organize photos or collect market data overnight. Personal uses sandboxed execution: installations and system-level changes occur inside cloud containers that are destroyed after a task, only explicitly authorized actions run on the user’s device, and heavy compute or higher-risk operations remain in DAPPOS’s cloud.

The platform provides a fast mode for simple daily tasks and a work mode that uses SOPs for more consistent results. Supported task types include voice dictation, text-to-speech, talking avatars, deep research, slide and document creation, fact checking, scheduled tasks, poster and image creation, video production and website development.

DAPPOS described xBubble as aimed at users who want deliverable results without learning model quirks, prompt structures or tool chains. The company plans to expand Engine’s ability to build SOPs for more complex tasks so a larger share of requests are handled by task-optimized execution paths, with the goal of improving output quality and reducing response times.

DAPPOS has raised more than $20 million from investors including Polychain, Binance Labs, Sequoia China, IDG Capital and OKX Ventures. In a release, DAPPOS stated that powerful AI should not require users to learn how AI operates and that xBubble uses AI to build and operate other AI solutions so users only need to state goals.