BashaShift

Your Next Home. Your Next Move. Skip the Chaos

Comment

BashaShift — Your Next Home. Your Next Move. Skip the Chaos.

Inspiration

Finding a home in Bangladesh is only the beginning. Renters often have to search for apartments, contact owners, schedule viewings, organize furniture, negotiate with movers, and coordinate moving day through disconnected conversations.

I wanted to solve this fragmented experience with one question:

What if finding your next home and moving into it could happen through one connected platform?

That idea became BashaShift, a mobile-friendly rental and relocation platform designed for Bangladesh.

What BashaShift Does

BashaShift brings three participants into one connected ecosystem.

For renters: Discover owner-published apartments, filter by area and budget, explore explainable property matches, request viewings, prepare a moving inventory, compare independent mover quotations, and confirm a booking.

For property owners: Publish rental properties with room photographs, manage listings, and respond to viewing requests.

For moving service providers: Discover eligible moving requests, submit independent quotations, and manage selected bookings.

The result is one connected journey:

Find a home → Request a viewing → Plan your move → Compare quotations → Confirm your mover.

Our AI Innovation: From Furniture Photos to Moving Plans

Moving preparation becomes easier with BashaShift's AI-assisted furniture recognition.

Renters can use their phone camera or upload furniture photographs. Using Mistral Small 3.1 Vision through Cloudflare Workers AI, the system analyzes images and suggests furniture items and quantities.

Users then review, correct, and approve the suggestions. The approved inventory contributes to an illustrative moving-budget calculation, together with distance, floors, and lift availability.

This is a human-in-the-loop AI workflow: AI assists with inventory recognition, while renters remain in control of the final information.

The budget uses transparent example assumptions rather than AI-generated market-price predictions. Actual moving providers submit their own quotations.

How I Built It

I developed BashaShift as a full-stack web application using:

  • React and Vite for a responsive, mobile-friendly user experience.
  • Node.js and Express for backend APIs and secure processing.
  • Supabase PostgreSQL, Authentication, Storage, and Row Level Security for user accounts, property data, photographs, and protected booking information.
  • Cloudflare Workers AI and Mistral Small 3.1 Vision for furniture recognition.
  • Render for hosting the public application.

The system supports three authenticated user roles and keeps customer contact details protected until a renter selects a moving provider.

Challenges and What I Learned

The biggest challenge was connecting three different user journeys while maintaining privacy, consistent booking states, and a simple user experience.

Integrating AI into production was another major learning experience. I encountered model licensing restrictions, API response-format issues, and deployment failures. Resolving these problems taught me that integrating AI involves much more than calling a model: it requires validation, secure image handling, reliable error management, and human review.

I also learned how to test role-based applications, protect database records through Row Level Security, and maintain a deployed full-stack project.

Validation and Current Progress

BashaShift is a publicly accessible working prototype.

The latest development suite reported 121 passing automated tests, including isolated acceptance scenarios with three renters, one property owner, and three moving providers. External authentication and AI responses were mocked in automated browser tests, while a real furniture-recognition result was separately demonstrated on the live application.

The platform currently uses user-published listings and independent moving providers. Their identities and availability are not independently verified, and no online payments are collected.

What's Next?

The next milestones are improving location-based moving assistance, validating more real-world rental and relocation scenarios, and piloting BashaShift with users in Dhaka.

My vision is simple: make the journey from finding a home to moving into it easier, more organized, and more accessible for people across Bangladesh.

Live App: https://bashashift.onrender.com

Source Code: https://github.com/IsratJahan973/BashaShift

Built With

Share this project:

Updates

Submission history