# Imohub: A real estate portal rebuilt at a fraction of the cost

> The portal had been patched for years, so it was rebuilt instead of patched again. It now indexes over 120,000 properties, answers queries in under half a second, and costs 70% less to run.

Sector: Real estate, Brazil
Delivered: before Velosyd existed, with the same approach, standards and tooling that run every Velosyd project.
Stack: Next.js, React, Laravel, MongoDB, AWS, Docker
URL: https://velosyd.com/cases/imohub-real-estate-portal

## Results
- **120k+**: Properties indexed and searchable.
- **Under 0.5s**: Query response across the whole index.
- **70%**: Infrastructure cost reduction.
- **Top 3**: Google rankings for the property searches it competes on.

## The situation
The portal had been patched for years. Search was slow enough to lose visitors, infrastructure cost more every month, and every new feature took longer than the last. The obvious answer was another round of patches.

## What was built
The honest call was that patching had run out. A rebuild is usually the wrong answer, so the decision was made against a test: could the existing search be made fast without replacing the storage layer? It couldn’t.

The rebuild put search on a properly modeled index, moved the front end to Next.js for rendering that search engines could read, and reshaped the infrastructure around what the traffic actually looked like.

## The result
The portal relaunched as Imohub, indexing over 120,000 properties with queries under half a second, 70% lower infrastructure cost, and top three Google rankings on its core terms.