BRENNEN

Industries / Construction and property / Real estate

Property management

Property management is running rented homes, shops and offices for the people who own them: letting them, collecting rent, keeping the accounts, arranging repairs, checking condition and meeting the law.

1864

John Ruskin bought three houses in Paradise Place, a London slum, and gave them to Octavia Hill to manage; the Octavia Hill Birthplace House says her methods "laid the foundations of the modern profession of housing management".

octaviahill.org

36

Milestones achieved across 6 domains.

10

To watch for: announced, not yet confirmed outside.

13

Open deliverables. Nobody has shipped them.

In one minute

Where it started

Octavia Hill managed Ruskin's houses with trained rent collectors who also watched over repairs and tenants, and her method spread to the Netherlands, Ireland and the United States. Over the next century law set the shape of the work: offices measured one way for rent from 1915, a repair duty written into short leases in England and Wales from 1961, and a US court ruling in 1970 that a lease promises a home fit to live in.

Where it stands

A 2021 survey by the US housing department and the Census Bureau counted 49.5 million rental units in 19.3 million properties, and found that about 84% of properties with 150 or more units were managed professionally, against about 22% of those with one to four. In England, letting agents registered about 70% of the 4.2 million tenancy deposits held in government-backed schemes in 2024. Software sellers now say AI agents answer tenants, settle repair requests and code invoices, and every such figure so far is the seller's own.

What is open

Most open jobs here are public tests that check what managers and their software do: whether an AI agent stops and sends help when a tenant reports a gas leak, whether screening pins a court record on the wrong person, whether federal inspection scores agree with a city's own violation records, and whether landlords' exemptions from England's energy rule hold up. The records needed are mostly public, and current AI models can read the leases, decisions and photos involved.

Domains

AchievedWatch forOpen

Figureheads

1 person

Advances to build with

1 research note

Open in property management

3 of 13 open deliverables

Open · Leasing

Spot unlawful lease clauses

A public test set of residential lease clauses, each labeled lawful or unlawful under one named jurisdiction, with the rule cited, and scores for AI models that review leases. Start with rules written as lists: England's Tenant Fees Act, which names the only payments a tenant can be charged, and Japan's 2020 Civil Code rules on deposits, normal wear and guarantees without a maximum. Write the clauses from public model leases and made-up variations, not from real tenants' leases.

Built with what exists

The statute texts on legislation.gov.uk and Japan's Ministry of Justice guidance; Label Studio for labeling; two reviewers with legal training per jurisdiction and a third to settle disagreements; current AI models as the reviewers under test; Hugging Face to host the set.

Credit · unclaimed

Open · Rent and accounting

Test service charge calculations

A public test of whether AI models and software can work out each tenant's share of a commercial building's running costs from the lease wording and the year's expense sheet. Write 200 made-up cases. Each holds lease clauses on the tenant's percentage share, the costs it excludes, a cap on yearly rises and "gross-up", which raises costs that vary with occupancy to what they would be in a fuller building, plus an expense sheet and a hand-checked correct bill for every tenant.

Built with what exists

Clause wording taken from office and retail leases filed as exhibits with the US Securities and Exchange Commission, found through EDGAR full-text search; Python to generate expense sheets and correct answers; current AI models with a spreadsheet or code tool; the free GitHub calculators as a baseline.

Credit · unclaimed

Open · Repairs

Safe self-fix test

A public test of the AI agents that walk tenants through a fix before a technician is sent. Write 200 scripted repair conversations. In half, a tenant can safely solve the problem: a running toilet, a smoke alarm chirping for a new battery, a thermostat with flat batteries, a blocked sink. In the other half the only right answer is to stop and send help at once, drawn from Victoria's legal list of urgent repairs: a gas leak, a dangerous electrical fault, a burst water service, flooding. A person or a second AI model plays the tenant from the script, and each agent is scored. No one touches real equipment; every conversation is simulated.

Built with what exists

The open maintenance-coordinator skill run in an agent harness; trial accounts of commercial maintenance agents; low-cost models such as Claude Haiku 5.5 or Meta's Muse Ember to play the tenant (see the Note "The jobs agents repeat"); a spreadsheet for scoring.

Credit · unclaimed

Where it sits