UnitPulse Platform

An AI-native leasing platform for property owners and their teams. It finds renters, converts them, and measures the funnel. We built it from zero starting February 2026, and the first paying customer signed a month later.

Leasing still runs on spreadsheets, 7 disconnected tools, and messages nobody answered. The platform is built around 3 people.

UP Marketing drafts posts and property content for each channel and each renter segment, and it keeps the Google Business Profile current, because “apartments near me” still lands there first.

UP Leasing turns that demand into signed leases. The AI answers every inquiry knowing who’s asking, writes follow-ups worth reading, and books the tour.

Agents drive all of it by typing. I first proved the Copilot pattern on a 2025 predecessor CRM, where it replaced a 15-click path with one sentence; here it fronts the platform’s whole AI suite.

UP Insight closes the loop for owners who aren’t analysts: the full funnel, per-prospect attribution, AI alerts, and a plain-language Data Copilot. Attribution shows which segments and channels actually sign leases, which then steers what Marketing writes next.

Customers came for Marketing, then asked us to run the whole journey. But the 3 products didn’t talk to each other — even though they consumed the same backend. 3 logins, 3 user-management systems, no shared front door for the AI. So we merged them behind one login and one design system, with the Copilot as the shared command layer — and rebuilt the IA so each persona sees what’s theirs without wading through everyone else’s. A phased merge, not a rewrite.

I led the UX layer and the design-system primitives; the absorption architecture was a shared effort with engineering. It’s built and merged today, but not yet cut over.

The 3 products were built fast, with AI writing Tailwind before AI was good at it — no maintained library, every screen its own local dialect, debt compounding quietly. Consolidating that was the real job: define the language once, build it as reusable components, then migrate the scattered UI onto it.

“One design language” is a claim, so here is the language itself. Warm cream surfaces, 2 greens with strict roles (olive for the brand, signal green for state — they never trade places), tabular numbers everywhere, 3 shadows in the entire system, dark mode built in. It lives as tokens and component CSS that the products build against, so the UI can’t drift apart again.

Dark mode is a token swap, not a second theme to maintain: the same component CSS reads flipped surface and state values.

I built this system in Claude Design, then extracted it to the tokens and component CSS the products compile against. The specimens above are rendered straight from that CSS, so what you see is what ships.

Leasing happens on foot, so we built the agents’ app from scratch in 2 weeks, and I decided what went in first. The feature I pushed hardest for was tour recording: it auto-summarizes the visit and sends the prospect a recap, so leasing’s highest-signal moment becomes data instead of a fading memory. Around it sit chat and calls, tour scheduling, prospect management, and the Copilot.

The sharpest evidence is one property that ran entirely on the merged platform: 93 units, 3 leasing agents, 2 months. Every one of these leads entered in that window, and every signed lease came from that same group, so it’s one funnel I can follow end to end.

Worth knowing how to read that last number: this is a new property in lease-up. We’re leasing all 93 units at once, including the ones a stabilized building would rarely list. Industry averages are mostly stabilized properties turning over a fraction of their units a year, so it isn’t a like-for-like comparison. It’s also why there’s no before-and-after here. There was no before.

The recordings are what made the drop-off legible. The same objections kept coming back: no parking, bedrooms with windows facing the stairwell or the elevator. I can’t say those cost us every lost lease, and I’m not claiming the platform caused a lift, there’s no clean baseline for that. What it did was run the whole funnel and show me where it leaked.

The portal is built and merged, but not yet cut over, and the convergence was a team effort I helped drive as design lead. What I’d measure next: funnel lift per segment, how many tours the AI books on its own, and whether the recap emails get opened.