Inspiration
The idea came from a shared experience: every member of our team has paid a late fee on rent at some point, without fully understanding the lease's late fee policy until the charge already hit. One of us also dealt with a water leak that made a room unusable for almost a month, only to be told the meant no compensation was owed.
Realizing this wasn't just one person's bad luck, but something we'd all been through, made it clear how common this problem really is. Most people sign leases, insurance policies, and other important documents without fully understanding the fees, deadlines, and protections buried in the fine print - until something goes wrong and it's too late to do anything about it.
That is why we built didyoureadthefine.ink! We created it to catch these problems before they happen: turning dense legal documents into clear red flags, verified concerns, deadlines, and next steps that anyone can actually understand, before they sign, not after.
What it does
didyoureadthefine.ink lets users upload any PDF document: leases, contracts, insurance policies or Terms of Service and get back an AI-powered analysis.
After uploading, the app extracts the document's text page-by-page and uses Gemini to identify:
Red flags: clauses worth reviewing closely, like non-refundable fees, automatic renewals, or one-sided termination rights - each backed by an exact quote and page number from the document Favorable terms: things that actually work in the user's favor Important dates and deadlines: lease start/end dates, notice periods, renewal windows Financial details: rent, fees, deposits, and other numbers to double check
The analysis is displayed as a highlighted, color-coded version of the extracted document, so users can see exactly where each finding came from. Highlighting it shows why it matters.
Beyond the one-time analysis, each uploaded document becomes a persistent "agent" the user can chat with. Since the agent has read the full document, it can answer follow-up questions grounded in the actual text, citing page numbers and refusing to guess when the answer isn't in the document.
Users sign in with Clerk, and their documents, analyses, and chat history are saved to MongoDB so they can come back and revisit past agents anytime.
How we built it
We built the frontend and backend together in a single Next.js project, styled with Tailwind CSS, and deployed on Vercel.
Clerk handles authentication, so every document is tied to a logged-in user and we never touch passwords ourselves.
PDF text is extracted in the browser page-by-page, then sent to Gemini, which does most of the heavy lifting:
Analyzes the document and returns structured red flags, favorable terms, key dates, and financial details - each backed by an exact quote and page number Names and classifies each saved document Powers a chatbot that answers follow-up questions grounded in the actual document text, citing pages and saying "needs confirmation" instead of guessing
MongoDB Atlas stores documents, analysis results, and chat history, scoped to each user.
The UI highlights AI-flagged passages directly in the document text, color-coded by severity, so users can see exactly why something was flagged.
What's next for Did You Read The Fine Ink ?
Deeper legal context - surfacing relevant tenant rights or state-specific regulations alongside flagged clauses, so users know not just what's unusual but what their actual rights are. \ Google Calendar integration - automatically add extracted deadlines (lease renewal dates, notice periods, payment due dates) directly to the user's calendar, so important dates don't get missed after the analysis is done.
Log in or sign up for Devpost to join the conversation.