How Mya works

Mya watches the tools your work lives in and turns raw activity into typed signals — a thread that needs a reply, a meeting with no agenda, a review request — so the same logic applies every time.

GmailOutlookSlackCalendarGitHubNotionJiraSalesforce

A named prioritization engine scores every signal on deadline, who it's from, and your priorities — then sorts it into clear tiers. The thresholds are explicit and configurable, not a black box.

Anything that touches the outside world passes you first. Mya proposes the move, you approve or edit, and only then does it act. You can override any call, set escalation rules, and undo within a window — the human stays accountable.

Replies

Rolando shared the updated document asking for final feedback

Reply← click reply

Team Updates

Your manager asked in #team-eng for an update on your work this week

Reply← click reply

Meeting Agendas

Your Design Review tomorrow doesn't have an agenda yet

Generate Agenda← click to generate

On your OK, Mya acts at the source — across tools, in sequence. Then it keeps the record current on its own: automatically tracking every project and updating its status — on track, at risk, blocked — as work moves.

Every project's status, tracked and updated automatically as work moves — no manual status checks.

Built for teams

Work is scoped, assignable, and accountable — across people and AI agents. Named roles govern access, every project has a visibility scope, and approvals route to the right person before anything ships.

Handoffs carry their full state, and approval chains route work across both people and AI agents (via MCP).

Memory — the company brain

Mya doesn't start from zero each session. It keeps a persistent model of your work — the state of every thread, project, decision, and blocker — layering what you tell it with what it learns.

That memory powers proprietary prioritization logic — a scoring engine with configurable rules and reusable playbooks that decide what rises to the top.