A personal operating layer
Afe connects the tools you already use, keeps a living timeline of what happened, and asks the smallest useful question when your judgment is the missing piece.
The PR merged, but the Jira ticket is still In Progress.
3 decisions from Slack and Calendar
“Incident review checklist” · 4 months ago
Bring the context you already have
The Afe loop
Afe is built around a simple loop: observe what changed, ask when your intent is unclear, and remember the answer with its source and scope.
Signals from Slack, GitHub, Gmail, Calendar, Jira, and your notes become a time-aware record—not another noisy feed.
Source events · timestamps · provenanceWhen the right next step depends on you, Afe asks a focused multiple-choice question instead of making a guess.
Small decisions · clear options · laterYour answers become scoped, inspectable events. Afe can resurface the reasoning when a future moment makes it useful.
Append-only history · corrections · outcomesAttention, with a reason
Afe gives every item a place and a reason. Urgent questions can interrupt. Normal work waits in Inbox. Useful context can rest in Shelf until it becomes relevant.
See the trust modelDeployment verification · due 14:00
Meeting brief · 3 decisions grouped
Incident patterns · resurfaced from notes
Recorded from your answer · project: payments-api
GitHub · linked to payments-api
Obsidian note · resurfaced because it is relevant
Memory that stays honest
Afe keeps observations, questions, answers, corrections, and outcomes as a timeline. You can follow the thread, see the evidence, export your history, or delete it.
Built for trust
Inspect, export, and delete the events Afe keeps about you. Nothing is hidden behind an “AI memory” label.
Material questions and recommendations show the source, timestamp, and relevant change before asking for trust.
Afe can prepare a safe action, but it cannot write to your tools without policy checks and your explicit approval.
Your context, everywhere
Afe is designed to be cloud-agnostic and multi-platform from the start: the interaction surface stays light while your context can scale in the cloud.
Early access
Tell us what you want Afe to notice first. We are building the product with knowledge workers who want fewer tabs, fewer interruptions, and better continuity.