RentRescue

Logo
HomePage
Upload
Analysis Results
Recommendations
Lease-Uh

Inspiration

Navigating the rental market in British Columbia is often a daunting and stressful experience for university students. Many students unknowingly sign leases containing illegal or unfair clauses that violate the BC Residential Tenancy Act (RTA). RentRescue was inspired by the need to empower renters with an accessible, AI-driven tool that helps them understand their rights and identify potential legal red flags before they commit to a contract.

What it does

RentRescue is an intelligent contract analysis platform that allows users to upload rental agreements in PDF or TXT format. Using Google's Gemini AI, the application performs a deep semantic analysis to detect:

High-Risk Violations: Excessive security deposits (exceeding 0.5 months' rent), non-refundable fees, and illegal eviction terms.
Medium-Risk Issues: Restrictive guest policies, improper rent increase terms, and pet deposit violations.
Informational Details: Clarity on utility responsibilities, notice periods, and subletting policies.

The tool provides a comprehensive risk score (0-100), color-coded risk levels, and specific actionable recommendations to help tenants negotiate terms or seek professional legal advice.

How we built it

The project utilizes a modern, privacy-focused tech stack:

Frontend: Built with React 18, TypeScript, and Vite for a fast and type-safe development experience.
UI/UX: Designed with Tailwind CSS and shadcn/ui components to ensure a responsive and professional interface.
AI Engine: Powered by Google Gemini 2.5 Flash for deep semantic understanding of legal text, handled through a dedicated service layer with custom prompt engineering.
Document Processing: Uses PDF.js for client-side text extraction, ensuring that contract data is processed in the browser.

Challenges we ran into

Parsing Complex PDFs: Extracting clean text from various document layouts, fonts, and structures proved difficult, requiring significant fine-tuning of the PDF.js implementation to maintain context.
Prompt Engineering for Legal Accuracy: Drafting prompts that consistently produce accurate legal interpretations of the BC RTA while adhering to a strict JSON schema was a meticulous process.
Handling API Latency: Ensuring a smooth user experience by providing instant feedback and visual cues while the AI processes the document.

Accomplishments that we're proud of

Seamless AI Integration: Successfully achieving deep semantic understanding of complex legal jargon using Gemini 2.5 Flash with minimal latency.
Privacy-First Architecture: Maintaining a high level of security by keeping document parsing and analysis results within the user's local session.
Digestible Insights: Developing a risk scoring system that translates complex legal concepts into a simple, color-coded format that is easy for any student to understand.

What we learned

BC Residential Tenancy Law: Gained a thorough understanding of the specific protections and regulations governing rental agreements in British Columbia.
AI Orchestration: Learned how to build robust AI-integrated applications with error handling, retry logic, and reliable fallback mechanisms.
Frontend Architecture: Improved skills in managing complex states for file processing and real-time analysis visualization.

What's next for RentRescue

Multi-Jurisdiction Support: Expanding the knowledge base to include rental laws and RTA regulations for other provinces and international regions.
OCR Support: Integrating Optical Character Recognition (OCR) to allow users to upload scanned images or photos of physical contracts.
Direct Legal Links: Connecting users directly to student legal aid services or the BC Residential Tenancy Branch for immediate follow-up.