What we are building, in the open.
What we are actively building, what is next in line, and what we are still exploring. Updated as scope changes.
Alert Action Center
Checklist of pending high-risk actions with done/dismiss workflow.
Useful / Noise feedback
Per-alert feedback that directly tunes detection thresholds.
Value-based paywall messaging
Upgrade prompts anchored in user outcomes, not feature limits.
Annual Pro plan (2 months free)
Stripe yearly option with clear savings badge.
Referral progress bar
Visible progress toward one free month of Pro.
Public roadmap and changelog
Weekly updates, fully public.
AI citation verifier
Block hallucinated quotes before alerts reach users.
Two-stage AI analysis
Cheap triage for most events, deeper analysis only on danger candidates.
Embedding-based semantic dedup
Per-host pattern memory to avoid re-analyzing identical rewording.
Feedback-driven threshold calibration
Nightly calibration of confidence floors from user feedback.
Nothing here yet.