Tire Swing launches an AI for affordable housing leasing and compliance: Reveneau

Why it mattersAffordable housing has rules every month. If an AI gets the paperwork right, qualified families move in faster. If it gets one rule wrong, the owner loses the subsidy for that unit.

Tire Swing launched on Y Combinator in the Fall 2026 batch. The product is an AI platform that handles leasing, outreach and eligibility checks for affordable housing. The YC launch page had 3 votes when we read it.

What it does

The platform reads a tenant's income documents, pay stubs and other forms, the company says. It checks the person against the programme's rules. It flags fraud. It routes the file to a staff member for sign-off. The company says a household can give full proof of eligibility in 30 minutes.

Tire Swing supports the main US affordable housing programmes, by its own account. The list includes LIHTC (Low-Income Housing Tax Credit) and PBRA (Project-Based Rental Assistance). It also includes Housing Choice Vouchers, HOME, public housing, USDA Rural Development 515 and HOTMA compliance.

The claim

The company says the manual coordination that the platform replaces costs operators more than 200 US dollars per transaction. It says the platform is live across dozens of properties in eight states, and runs for multiple funding programmes. Neither figure is checked by an independent party.

The founders and the funding

The founders are Paul Witten and Lucas, both computer science graduates from Princeton, per the YC launch page. The page says Paul led cybersecurity investigations at a previous employer. Lucas was a senior software engineer at Instacart, per the same page. The company does not publish a funding amount.

The company site shows a product tour built around a fictional dataset, Maple Court Housing: 348 units across five properties in Ohio and Illinois. The site states this is a sample, not a real customer.

The hard part

Affordable housing rules change by programme, by state and often by property. An error is expensive: a wrong income count can put a household in the wrong unit. The owner can then lose the tax credit for that unit. A recertification missed by the deadline can cost months of subsidy. So an AI that reads the file has to be right the first time, every time.

The company's own line is that staff still sign off on each file. That is the right stance for a regulated product. A buyer should ask how often the AI's recommendation needs a correction. A buyer should also ask in which part of the file those corrections sit. Published numbers would make the product easier to adopt.

Source

This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.