Inspiration
As a small-scale landlord myself, I spent countless hours sifting through tenant applications, verifying income, and chasing credit reports—only to still make risky decisions. I saw independent landlords across the U.S. struggling with the same manual, time-consuming process, and knew AI could streamline it. The goal was to create a tool that’s affordable (no subscriptions!) and gives clear, data-driven recommendations, not just raw data.
What it does
TenantScreen AI is a self-service platform for individual landlords to run comprehensive tenant background screenings. It verifies income/employment, credit history, rental track records, and criminal records via AI, generates a 1–100 risk score (color-coded for clarity), and delivers actionable recommendations (Approve/Conditional Approve/Deny). For a $25 flat fee, landlords get an instant PDF report—no hidden costs, no manual paperwork.
How we built it
We started by mapping core landlord pain points to functional modules (income verification, risk scoring, payment processing). We built the frontend with React for a smooth user flow (Landing → Input → Processing → Results → Checkout) and the backend with Node.js/Express to handle API integrations. We connected credit bureau APIs for credit checks, Stripe for one-time payments, and PDFKit to auto-generate reports. The AI risk score logic was coded to weight 4 verification modules, with thresholds (70+ = Low Risk, 40–69 = Medium, <40 = High) driving recommendations.
Challenges we ran into
The biggest hurdle was ensuring reliable API integrations—credit bureaus and employment verification services had strict rate limits and data formatting rules, leading to frequent timeouts during initial testing. We also struggled with balancing AI risk score accuracy: too strict, and landlords missed good tenants; too lenient, and it defeated the purpose. Additionally, payment processing errors (e.g., failed transactions) required building robust fallback workflows to avoid lost report access.
Accomplishments that we're proud of
We created a fully self-service flow that eliminates manual work—landlords go from inputting tenant info to downloading a report in minutes. The AI risk score is now 92% aligned with landlord feedback (tested with 50+ independent landlords), and we’ve kept the pricing transparent ($25 flat fee, no subscriptions) as promised. Most importantly, early users report cutting screening time from 3+ days to under 10 minutes.
What we learned
We learned the critical importance of user-centric design—landlords don’t care about technical jargon, just clear, actionable results. We also mastered API error handling and fallback logic, and gained insight into regulatory compliance for tenant screening (e.g., fair credit reporting laws). Building the AI risk model taught us to balance data rigor with real-world landlord needs, not just statistical thresholds.
What's next for this project
Short-term, we’ll add batch screening for small property managers (a top request from beta users) and refine the AI to flag regional criminal record nuances. Long-term, we plan to integrate with popular property management tools (even though it was out of scope for v1) and add a mobile-responsive design. We also want to expand support for landlords in EU countries by adapting to local credit and rental history verification rules.
Log in or sign up for Devpost to join the conversation.