Drag a Client onto a Property Manager to re-parent. Click an org to expand / edit.

Manage individual sign-ins, roles, and invites for this organization.

Platform-wide org management — create, edit, suspend, or delete organizations, manage billing, and invite or edit their users. Click an organization to expand.

🟢 online (active in the last 3 min) · ⚪ logged in but idle. Refreshes every 30s while open.

Ranked by real time spent using the app (measured from activity heartbeats).

When ON, everyone except superadmins sees a full-screen Scheduled Maintenance page. Set a message and an expected-back time so users know what to expect and when to return. Message/ETA changes take effect on each user's next page load.

Emails superadmins once a day when a property's available-unit count drops sharply. The two windows fire independently — a cliff overnight alerts even if the week looks flat, and a slow bleed across a week alerts even if no single night moved much. Counts use the same availability rule as the Property Summary tiles, so an alert can be checked against what you see in the app. No email is sent when nothing trips a threshold.

Set owner / management / payer org per subject property.

Cross-checks each property across three sources — our own-site scrape, apartments.com, and Zillow. Surfaces agreements (high confidence), conflicts (sources disagree), gaps (only one source has it), and suspect values. Report-only — nothing is written to live data.

Checks that every physical unit carries one consistent identifier (building#-unit# where a building exists) across every scrape. A clean property needs no attention; an inconsistent one stores the same unit under multiple spellings (e.g. both 1112 and N/A-1112), which change-detection would read as phantom churn. Report-only.

Live VM resource usage (RAM, CPU, swap, disk) and the residential WireGuard proxy fleet used for rendering. Flip a proxy's In rotation switch to include or exclude it from the scraper's selection.

Scheduled jobs run in the nrdata-scheduler process (all times UTC). Next-run is computed from the schedule; last-run & in-progress come from the run log. Run now fires a job immediately — heavy jobs (scrapes, ranking captures) add real render load.

One row per scrape run: how long it took, how many properties it reached, and the net change in available units. Click a run to see every property in it — sort by Δ Units to find what gained or lost inventory. A property with no earlier recorded run shows “—”, not 0: a missing reference is not a zero change.

Tracks the last successful unit-write per subject and per comp. Subjects that miss the nightly window (e.g. Pegasus two-step click sites that exceed the 480s per-property budget) surface here as "very stale" or "critical" before anyone notices in the dashboard.

Manage which comps appear under which subject, and (for superadmins) which organizations have adopted which properties. Move a comp by picking a new subject from the dropdown; remove a link to unsubscribe the org from that comp.

Per-property tools are native on each Scrape Health row above: ✎ Edit for name / units / management / facts, and ⚙ Source to set the scrape source (platform + API URL, secondary source, Knock community URL). Use Auto-detect API URLs to bulk-register knock/sightmap sources for properties that don't have one yet.

The live Stripe account the app bills through. Keys are stored in Google Secret Manager and never shown here. Paste a new value to rotate it — the key is validated against Stripe before it's saved, and takes effect within ~a minute (no redeploy). The secret key and the price IDs must belong to the same account.

Each subject property gets its own billing row when added (auto-created based on the org's default-payer setting). Active = paid, Unclaimed = needs a payer, Grandfathered = MKTSVY-paid (no charge). Use "Release payer" to force a paying subject back to unclaimed so a different user can claim it.

This organization's plan, usage vs. limits, and Stripe Customer Portal access for invoices & payment methods.

Each widget has two global controls on the right — the visibility checkbox (shows/hides the feature for all users; off removes it from the app) and a clickable Beta pill (colored = shown, faded = off). On the left, the default-layout controls — drag order, the "visible by default" checkbox, and width — set the STARTING dashboard for brand-new users, resets, and the demo; existing users keep their own arrangement.

Demo properties

Choose which subject properties appear in the public Live Demo (/app/v2?demo=1). Checked properties are shown live — with their comps, charts, and downloadable reports; unchecked are hidden. Saving takes effect immediately and clears the demo report cache.

Report delivery

Controls how the public Live Demo (/app/v2?demo=1) serves PDF/Excel reports for the sample portfolio. Live generation renders every report fresh — always current, heavier on the server. Cached serves a stored copy that auto-refreshes every 24h — switch to this if the demo draws heavy traffic. Only the demo is affected; real accounts always generate live.

Captured leads

Prospects who used “Email me a report” in the demo — no account required. Each row is a real, consented email address you can follow up with or import into your CRM.