Pulse

The dashboard leads with one instrument: portfolio occupancy, weighted review rating, open threats, and comps-cut-rent — with deltas, a trend sparkline, the live action feed, and the market-context row (rate climate, rent vs area, affordability, jobs). Below it: the geographic map with per-community health pins and the community grid. Numbers are live; when something has no data yet, the instrument says so instead of faking it.

Each community has a tracked comp set. Sentinel captures competitor pricing on a schedule — direct scrape first, AI web-research fallback — and keeps full price history. The moves feed surfaces changes ("Mirabella ↑ $40 on 2BR · added 1 mo free") and the AI brief summarizes the week's competitive picture per community.

For every community: where do we sit on the local price ladder? 1 = cheapest, N = most premium. Each card shows the rank badge, the comp-range bar with your position dot, and the actual dollars (us / comp range / comp average) for 1BR and 2BR. The page header counts your portfolio's positioning at a glance: cheapest-in-set, value half, premium half, data-thin.

✦ Ranking needs 2+ active comps with current pricing. "Data thin" cards tell you which comp sets to grow.

All Google reviews across the portfolio in one triage queue, priority-ordered: unresponded 1-2★ first (P1 with an SLA clock), then P2/P3. Bulk triage, AI-drafted replies grounded in verified property facts, a "Replied on Google" tracking flag, and a network rating trend sparkline from your Places history. Other rating sources appear alongside for context.

Free public data, pulled with your own keys: Census ACS (area rent + income), HUD Fair Market Rents, BLS employment, FRED rates. Sentinel stores it per state AND per community ZIP (ZIP-precise where pulled), shows each community's rent vs its area median and the 30%-of-income affordability ceiling, and feeds the same numbers to the Pricing Advisor, campaign briefs, and the CRM renewal radar. Sources are labeled on every surface.