DANDA #016 — RentRelief: AI Agents to Stop Evictions Before They Start

Millions of Americans are forced from their homes every year—not because they can’t pay, but because they’re overwhelmed by confusing paperwork, deadlines, or simply don’t know their rights. As rents rise and affordable housing shrinks, the eviction pipeline is surging. Landlords often act fast; tenants miss a letter or a deadline, and suddenly, a family’s out on the street. Today, AI can step in as a tireless, impartial agent—taking real, practical action before eviction happens.

The problem

Eviction is a routine disaster for U.S. renters. In 2024, over 3.6 million eviction filings occurred nationwide. Many tenants lose their homes not because of inability to pay, but because of missed court appearances, incomplete paperwork, or a lack of legal know-how. The process is quick: in some U.S. cities, you have as little as 5 days to respond to an eviction notice. Legal aid is scarce—less than 10% of tenants have legal representation versus 81% of landlords. The downstream effects hit hard: evicted families are 11 times more likely to enter a homeless shelter, and an eviction record can block future rental opportunities and even job prospects. Billions in government rent relief funds go unused every year due to application complexity or missed deadlines. The pain is preventable, but the system is stacked against renters.

The idea: RentRelief

RentRelief is a fully agentic AI co-pilot that intervenes at the first sign of eviction risk—monitoring communication, automating applications for relief funds, drafting personalized legal responses, and negotiating directly with landlords. The agent acts as a virtual “tenant advocate,” alerting users to deadlines, generating all required paperwork, and even connecting with legal aid or social services. By embedding itself in the renter’s digital life—email, text, rent portals—it proactively flags risk and takes action, closing the gap that gets families evicted for technicalities or lack of knowledge.

Architecture

Here’s how RentRelief works: Inputs like emails, texts, and rent portal alerts feed into the Ingestion layer, where documents are scanned, parsed, and key entities (dates, landlord name, court info) are extracted. Memory/Graph builds a real-time timeline and stores all documents, deadlines, and user data. The Agent Orchestrator (powered by LLMs and legal rules) analyzes for eviction risk, generates personalized responses, and auto-fills relief forms. For critical actions—like contesting in court or applying for aid—the Human Gate checks with the tenant or legal aid. The Action Layer submits filings, notifies users, and negotiates with landlords automatically, closing the loop.

Build plan (90 days)

Wedge: Target renters in the 10 U.S. cities with the highest eviction rates and most active rental portals. Build integrations for popular rent payment/email systems and automate state-specific legal forms.
Stack: Expand support to cover 80%+ of U.S. legal jurisdictions, partner with legal aid orgs, and introduce AI chat for deeper negotiation and dispute mediation.
Pricing: Direct-to-tenant $19/mo (free for users referred by legal aid or social services), with a SaaS model for non-profits and city governments ($2/user/mo at scale).

Why now

Eviction tech is stuck in the 1990s, while AI can now ingest documents, reason across legal timelines, and personalize action in seconds. Courts are digitizing filings, and legal aid APIs are emerging. Renters face record-high incomes-to-rent ratios and more digital landlord comms than ever before. RentRelief brings real agency to the people AI forgot—precisely when housing insecurity’s never been higher.