Hously

Challenge

Canadian buyers and sellers have one credible portal and a handful of brittle alternatives. The brief was twofold: build a consumer-facing portal that beats Realtor.ca on sold prices, neighborhood data, and calculators that respect Canadian rules (CMHC, LTT, NRST, vacant home tax, FHSA), and pair it with a brokerage-side CRM whose lead intake, campaigns, and reporting all run off the same SSOT — no parallel tables, no hidden state.

Approach

1. 01

   Designed a phased twelve-month roadmap (foundation, differentiation, scale) so each phase shipped an anchor metric and a defensible piece of the data moat.

2. 02

   Built listing infrastructure as an event stream: every list, relist, price change, terminated event captured for downstream sold-price analytics, AVM training, and saved-search alerts.

3. 03

   Stood up the CRM as a strictly database-driven SSOT with dynamic pipeline stages, lookup catalogs, and stage automation rules — no business records hardcoded in UI flows.

4. 04

   Built an in-house cron worker rather than relying on platform cron, dispatching campaigns and deadline alerts with Resend and Twilio adapters and full webhook delivery telemetry.

5. 05

   Wired Sentry across server, edge, and client runtimes; gated CI on migration checks, type checks, and builds; shipped a PWA-installable client.

Outcome

The consumer portal and brokerage CRM share one schema, one event stream, and one campaign engine. Listings, calculators, AVM, saved searches, and brokerage campaigns are not separate products — they are projections of the same source of truth. The roadmap is sequenced for the platform to become the Canadian source for sold-price and neighborhood data.