RentProof

Inspiration

Securing a student rental is a financial minefield of hidden fees and bait-and-switch leases. Our motto is simple: "Talk is cheap. Show me the lease." We built RentProof to empower renters to uncover red flags, hold landlords accountable to their original promises, and sign contracts with total confidence.

What it does

How we built it

Frontend & State: A highly responsive Next.js application built to handle complex "what-if" financial scenarios and private comparisons.
AI Extraction: Gemini (with a fallback model) pulls exact financial and legal data from text and PDFs.
Research Pipeline: Integrated Tavily search—strictly limited to team-reviewed sources—alongside municipal APIs.
iMessage Agent: Leveraged Photon cloud numbers to connect our backend logic with native iOS messaging.

Challenges we ran into

AI hallucinations in legal and financial contexts can lead to expensive mistakes. Forcing our extraction models to map every single data point back to an exact, verifiable source quote was a massive architectural hurdle. Additionally, normalizing unstructured data from diverse listing formats and designing a clean SMS interface for complex data required intense problem-solving.

Accomplishments that we're proud of

We built a real-time consumer protection tool. Giving users the power to automatically catch a discrepancy between an online listing and a physical lease is a huge win for transparency. We are also proud that users remain in total control. If the AI misses a detail, users can manually correct listings with a full, trackable version history.

What we learned

We learned how to successfully ground LLM outputs by enforcing strict source-linking, ensuring users never have to blindly trust the AI's math. We also mastered complex state management to support manual overrides and discovered how to distill dense municipal data into actionable insights for renters.

What's next for RentProof

We are finalizing our image and scanned-PDF extraction so users can instantly analyze paper leases with their phone cameras. From there, we plan to expand our building research integrations to student housing markets beyond NYC and refine our iMessage agent to make group room-hunting completely frictionless.