LendIt

Hyper-local tool sharing platform with AI-based trust scoring.

Comment

LendIt

Inspiration

LendIt started from a simple frustration we kept seeing around us: people buy expensive tools for one-time use, while the same tools often sit unused in someone else’s home.

We wanted to build a trust-first local borrowing platform that makes it easier to share tools within nearby communities instead of purchasing new ones every time.

We were especially inspired by two questions:

  1. How can we make local borrowing feel as reliable as buying?
  2. How can we reduce waste and unnecessary spending with a better user experience?

What it does

LendIt is a hyper-local peer-to-peer tool sharing platform that allows users to:

  • List the tools they own for others to borrow
  • Find tools available within a 10 km radius
  • Book tools easily through a simple interface
  • Leave reviews after transactions to build trust

A key feature of LendIt is its AI-based risk scoring system.

  • Supports ML-based inference when available
  • Uses rule-based fallback scoring when ML is unavailable

This ensures reliability while maintaining fairness.

How we built it

Frontend:

  • React + Vite application for authentication, tool discovery, booking flows, review submission, and profile navigation

Backend:

  • Spring Boot (Java 21) REST API with JWT authentication, role-based booking logic, validation, and business rules

Database:

  • MySQL with JPA/Hibernate for managing users, tools, bookings, reviews, categories, blocked dates, and risk scores

Trust Engine:

  • Python-based ML inference pipeline
  • Includes fallback scoring system for reliability

Location Intelligence:

  • Client geolocation + backend Haversine distance calculation for nearby tools

Automation:

  • Scheduler that auto-completes expired bookings

Challenges we ran into

  • Designing booking workflows: Managing state transitions and permissions between owners and borrowers
  • Preventing overlaps: Handling date conflicts and race conditions
  • Fair review system: Ensuring one review per booking and correct user roles
  • ML integration reliability: Supporting ML with fallback to avoid system failure
  • Location UX issues: Handling denied permissions smoothly
  • Backward compatibility: Supporting evolving features without breaking older data

Accomplishments that we're proud of

  • Built a complete end-to-end platform from UI to backend and trust scoring
  • Developed a trust pipeline combining reviews with risk classification
  • Implemented resilient architecture with ML + fallback logic
  • Handled real-world constraints like blocked dates and booking conflicts
  • Created a smooth user experience for both borrowers and owners
  • Designed a scalable foundation for future growth

What we learned

  • Trust systems go beyond algorithms and require strong product design
  • Fallback design is critical for system reliability
  • Booking systems need strong state modeling to avoid edge cases
  • Geolocation requires clear UX handling for permissions
  • Full-stack integration requires tight coordination between frontend and backend
  • Real workflows reveal hidden complexity faster than isolated development

What's next for LendIt

  • Add payment integration
  • Improve AI risk model accuracy
  • Introduce real-time chat system
  • Enhance location intelligence
  • Deploy for real-world community use

Built With

Share this project:

Updates