Rentout - Paris Vibe-a-thon: “Build a One-Person Unicorn” — November 15, 2025
AI Tinkerers - Paris
Hackathon Showcase Finalist

Rentout

Team consisting of Arihant (SDE II, Thriwe; EPITA MSc; Go/Python), Paula (Founder & CEO, Oly; MSc X‑HEC), and Anastasios (Talos AI; Maastricht) — backend, AI commerce.

3 members Watch Demo

Project Description — RentOut

log-in: [email protected]
password: 1234567890
https://rentout-ai-host.lovable.app/

RentOut is an AI co-host that fully manages short-term rentals autonomously, solving the operational overload faced by small hosts who manage multiple listings across different platforms. Hosts aged **35–55 with at least two rental units—people who treat short-term rentals as a side revenue stream—struggle with manual availability syncing, competitive pricing, seasonal adjustments, guest communication, cleaning coordination, and gathering reviews. RentOut eliminates these manual tasks by acting as a self-running property manager, delivering consistent execution, optimized occupancy, and a seamless guest experience.

RentOut centralizes and automates the entire workflow: cross-platform calendar synchronization, dynamic pricing analysis, automated guest messaging, post-stay review follow-ups, and maintenance/cleaning orchestration. Hosts simply set their goals—maximize occupancy or protect minimum nightly revenue—and RentOut handles everything else. By combining autonomous agents, marketplace data, and multi-platform integrations, RentOut provides the reliability of a full property-management service without the cost or complexity.


How We Meet the Judging Criteria

Execution & Functionality

RentOut demonstrates a complete end-to-end operational loop: listing management, pricing, messaging, booking handling, issue resolution, and cleaning scheduling. The demo shows real-world scenarios, agent interactions, and system reliability.

AI & Agents Usage

The system uses autonomous agents for:

  • Calendar reconciliation
  • Pricing optimization and competitive analysis
  • Guest communications with context and memory
  • Incident management (e.g., damage, misplaced items)
  • Cleaning/maintenance task routing
  • Review follow-up nudges
    Agents operate with chain-of-thought suppression, tool usage, and safety-aware fallback behavior.

Scalability — One-Person Model

A single operator can manage hundreds of listings because the system auto-executes 90%+ of routine operations. All workflows are designed for “fire-and-forget” use: once a listing is connected, RentOut handles daily tasks automatically.

Problem Clarity & Market Impact

Short-term rental hosts are drowning in fragmented tools, unpredictable pricing, platform-specific inboxes, and maintenance chaos. RentOut integrates these into one platform and autonomates the entire tenant lifecycle, enabling small hosts to compete with professional property managers and maintain high service standards.

Demo & Product Narrative (Video Flow)

  1. Host with three apartments connects Airbnb/Booking calendars via RentOut.
  2. System performs cross-listing and availability sync.
  3. Host selects an objective: maximize occupancy or respect a minimum nightly price.
  4. A guest inquiry arrives — agent replies automatically and confirms booking.
  5. During the stay, guest reports “a broken cup” — RentOut handles the situation, logs damage, and arranges replacements.
  6. Guest asks: “Where can I buy fresh fish nearby?” — AI replies with a relevant city recommendation.
  7. After checkout, guest receives an automated review request with a personalized message.
  8. RentOut schedules cleaning service before the next check-in, pulling in maintenance providers integrated on the platform.

This linear story demonstrates the product’s full functionality and autonomy.

Partner Tech Usage

RentOut leverages marketplace APIs (Airbnb/Booking), LLMs, embeddings, vector memory, and structured tool calls. Cleaning and service providers are integrated through partner webhooks or APIs.


Technologies Used

AI & Backend:

  • Claude 2.1, HuggingFace, OpenAI GPT-4o-Mini, Vertex AI; Default: Claude 2.1
  • Temporal / Celery for task orchestration
  • Python FastAPI backend
  • Redis for queues + caching
  • Postgres for structured property + booking data
  • N8N.io

Integrations & APIs:

  • Airbnb or Booking.com calendar APIs
  • Google Places (for local recommendations)
  • Stripe for payments/incident charges
  • Twilio / WhatsApp / email messaging channels
  • Cleaning/maintenance provider APIs or custom webhook endpoints

Frontend & Hosting:

  • Lovable & Lovable Cloud

Datasets:

  • In this demo we’re mocking data. However given more time we would use the datasets provided by the different marketplaces API’s, such as Booking.com ( https://developers.booking.com/connectivity/docs/ari )

Autonomous Agent Logic

  • Calendar Agent: syncs availability across platforms; resolves conflicts.
  • Pricing Agent: runs competitive analysis, seasonal curves, demand estimation; enforces user constraints.
  • Guest Communication Agent: handles FAQs, arrival info, neighborhood suggestions, and issue escalation.
  • Ops Agent: coordinates cleaning, sends tasks, checks status.
  • Review Agent: messages guests post-stay and optimizes language to drive reviews.

Failure Modes & Fallbacks

  • Conflicting availability → human review queue
  • Ambiguous guest issues → escalate to host with proposed solutions
  • Pricing anomalies → revert to safe minimum nightly rate
  • Integration outage → local queue retries with exponential backoff

Next Steps

  • Add automated photo analysis to detect damage after guests leave.
  • Launch revenue-prediction dashboard for hosts.
  • Expand integrations to more platforms : Expedia and local cleaning networks.
  • Introduce autonomous “Smart Renovation Suggestions” using trends + sentiment data.
Google Lovable n8n