Get Inspired

Sometimes it can be hard to find the perfect Zuhause, property, or land. The process of looking can be quite extensive, time consuming, and involve many trips. When looking to help people save time finding their ideal home, land or property, I designed a user-friendly map-based platform. Using the Gemini 3 AI Engine to enable users to search by area, drag the map to find cattle/horses/not-fish; finding their ideal home will be much quicker and easier.

How does it work?

LandVerify AI uses a three-stage AI process to help users identify and verify properties. The First Stage uses identity verification and is performed through document scanning. The Second stages process involves a document comparison to ensure the identity document matches the scanned image; and the last stage will assist in conducting a walk-through view of the actual property through a video confirmation.

How we built it The source code for our application was built using Next.js with React & Tailwind for the front-end, back-end using Node.js (with Prisma), PostgreSQL (with PostgreSQL), Clerk (user authentication), We built a map using Leaflet maps, and used the Gemini 3 Flash as our main AI engine.

Obstacles We Encountered

Different types of document formats, getting video processed quickly and numerous, AI modules processed together smoothly (the entire process must flow seamlessly).

What We Accomplished

LandVerify AI provides a complete, three-stage: Identity verification; Document Verification; Video verification process that allows users to easily find properties using a map.

Lessons Learned

How to develop multi-modal AI processes to connect and create real user experiences.

What’s Next for LandVerify AI?

Improve search function; personalising based on previous searches; Real-time property verification.

Built With

  • clerk.
  • gemini3flash
  • nextjs
  • node.js.
  • openstreetmap.
  • postgresql
  • tailwind
Share this project:

Updates