Find the leak. Quantify it.
Dwelleroo compares your portfolio against its own history to surface where money is quietly going. Dwelby explains the findings, drafts the monthly review, and cites the numbers it used.
Two units account for 74% of current value at stake. Resolve the Birch House turn, then review the repeated HVAC pattern.
Dashboards tell you what happened. They rarely tell you what it cost.
- 01A unit takes two weeks longer to turn than it used to, and nothing announces it — you just notice the year was worse.
- 02Emergency call-outs creep up gradually, so no single invoice ever looks alarming enough to investigate.
- 03Arrears age quietly in a column nobody sorts, until the balance is too large to recover politely.
- 04The numbers exist somewhere in the reports, but assembling them into a decision is a job nobody has time for.
Patterns, priced, in priority order.
Portfolio pulse
A standing view of value at stake, value realized, and findings resolved — so the picture is current rather than assembled once a quarter.
Money-leak detection
Vacancy loss, arrears aging, expense creep, emergency-work ratios, and unusual unit-turn costs, each quantified in dollars per year.
Monthly portfolio review
A written narrative built from server-computed aggregates, citing the figures behind every claim so you can check the work.
Value ledger
Tracks what actually landed — late balances collected, vacancy closed — kept separate from estimates so savings are never double-counted.
Dwelby can explain and draft. It cannot quietly change your records.
Plenty of software is racing to let a model act on your behalf. We think a system that holds tenant ledgers and lease terms should earn that in much smaller steps — so model calls do not write to core business tables. Dwelby prepares the action; a person commits it.
Read-only model access
Dwelby explains, drafts, extracts, and prioritizes. It does not write to the tables that run your business.
Deterministic findings
The money-leak numbers are computed on the server from your aggregates. The model narrates them; it does not invent them.
Your baselines, not the market's
Comparisons are against your own portfolio history, not pooled rent data from properties you have never seen.
Estimates stay separate
Value at stake and realized value are tracked as two different figures, so nothing inflates a counterfactual saving.
It only works because the other five do too.
Each area writes to the same records, so nothing needs copying between them.