What it is
Gentle Landlord is a rental-management app for small landlords, built as a prototype for a friend's idea. Tenants send messages and maintenance requests through a portal. The landlord sees a dashboard of what needs attention. For each message, an AI is meant to write a short briefing and a draft reply in the landlord's chosen tone, which the landlord reviews before sending. The idea is that the AI acts as a buffer: it holds to the landlord's own policies while keeping the language kind.
How it works
- A tenant logs in to their portal and sends a message or a maintenance request.
- It shows up in the landlord's attention queue.
- The landlord opens it and sees a briefing card plus a drafted reply.
- The landlord edits or approves the draft and sends it.
- Maintenance, change requests, payments and tenant records each have their own page.
What exists today
- Owner dashboard pages: attention, maintenance, modification requests, messages, payments, tenants, settings
- A conversation-first tenant hub showing the AI draft
- A tenant portal for maintenance and messages
- A pretend login so you can click through as owner or tenant
- A database schema and sample data
- Canned AI replies that show the intended flow
- Browser tests that click through the pages
What isn't there yet
- Every page runs on sample data; there is no real backend
- No real AI calls yet; replies are scripted
- No real login, payments, text messages or email
- Planned pieces not started: maintenance calendar, eviction case tracking, hosting
- Work paused after the owner dashboard and tenant hub were built
Under the hood
- Next.js and React for the app
- Prisma with Postgres for the data model, seeded with sample records
- One mock-data file drives every page, and one mock module stands in for the AI
- Playwright tests run in a real browser