Journey
From a personal dashboard to an operating system for AI work.
amoOS grew by closing one gap at a time: the gap between having useful AI tools and being able to trust them with longer, connected work.
Evolution
Personal workspace
Projects, tasks, knowledge and chat, unified in one place.
→ One place to think
A persistent Brain
Memory went from raw retrieval to a compiled wiki with provenance, review and evaluation.
→ Continuity
Reachable from anywhere
MCP and Telegram made the same system available from any AI client and any device.
→ Presence
The fleet
Cross-machine dispatch, project-aware workers, isolated worktrees and a full session lifecycle.
→ Scale
The command center
Live sessions, activity, system health, usage and analytics made the fleet observable.
→ Visibility
Learning loops
Handoffs, lessons, verify passes and definition-of-done gates turned every run into input for the next.
→ Compounding
Voice and devices
A native voice app, and a device plane that lets any agent drive phones and desktops.
→ New surfaces
Longer autonomy
Per-project model accounts, usage-aware pacing and fewer, better interruptions.
→ Trust
Lessons
Autonomy is a trust problem
Memory must change behaviour
Telemetry is a control surface
The north star
Less time supervising routine execution. More time on direction, judgment and people, while the system keeps everything visible and compounds what it learns.