Meet the secure personal AI agent you can trust.
What should X1 take off your plate?
X1 is a personal AI agent by XScaleAI. It installs in hardened Docker containers on your own machine, so your files, your notes, and the work it does for you stay on hardware you control. You choose one job worth repeating, X1 runs it on your schedule, and every result is reviewable before you expand its responsibilities.
What X1 can do
- Keep your job search moving. Keep roles, company research, application notes, follow-ups, and next steps together so you always know where to pick up.
- Reach a decision faster. Bring credible sources into one brief that separates agreement, uncertainty, and the questions still worth checking.
- Know what changed in your market. Review direction, risk, sector leadership, and watchlist changes in one focused view before deciding what deserves attention. Market watch is powered by Signl.
- Watch a topic without living in feeds. Collect relevant changes around a regulation, competitor, security issue, or industry theme for a calmer review cycle.
Start useful. Expand deliberately.
X1 is not a promise to automate your whole life. It is a practical way to take one repeated job, make the output reviewable, and build from proof.
- Choose one job worth repeating. Define the outcome, timing, and sources you care about so X1 starts with a narrow, understandable responsibility.
- Let guided setup handle the plumbing. X1 packages the agent runtime and configuration so setup does not begin with a weekend of Docker troubleshooting.
- Review the result before you expand. Check the sources, correct the preferences, and decide whether the first job has earned enough trust to take on another.
How X1 keeps your work private
X1 runs locally in containers that are read-only, drop all Linux capabilities, cannot gain new privileges, and are never given access to the Docker socket. The portal binds to your loopback interface, so it is not exposed to your network. Notes and results are written to files on your own disk.
Cloud models are the default route for the reasoning itself, through OpenRouter, so the context a task needs does leave your machine under that provider's terms. Ollama is the local option if you would rather keep inference on your own hardware too. X1 tells you which connections are active before you turn them on.
Getting X1
X1 is in open beta and free to use during the beta. It needs Docker and your own OpenRouter API key. Installation is a single command on macOS, Linux, and Windows.