Skip to content
cd ../projects
[ CASE_STUDY // 01 ]

Reisearch Platform

An entire real estate investment platform architected and shipped end to end.

React 19TypeScriptGoPythonAWSDynamoDBServerless
Reisearch Platform — main interface screenshot
34FE Modules
16Go Services
9.6MRecords
−40%Infra Cost

Project Overview

Reisearch is a complete real estate investment platform built by a small engineering team. Investors use it to find comparable properties, study local markets, model deal economics, and visualize what a property could become after renovation.

By the Numbers

  • 34 feature modules in the React 19 frontend.
  • 16 Go microservices behind the platform API.
  • 23+ DynamoDB tables designed for the access patterns the product actually needs.
  • 9.6M housing market records ingested into queryable form.
  • 40% reduction in infrastructure costs after the AWS architecture redesign.
  • 30% reduction in production bugs after introducing strict typing, hypothesis-based tests, and pre-commit gates.
  • 25% increase in user engagement after the redesign.

My Ownership Inside the Platform

Several systems on the platform are mine end to end. Designed, shipped, maintained:

  • Comps Orchestration Pipeline (Go): the 40-minute to 4-minute throughput rebuild.
  • AI Renovation Copilot (Python / FastAPI): OpenAI plus Gemini, real-time streaming, itemized cost estimates.
  • Market Data Pipeline: 9.6M housing records normalized into queryable DynamoDB tables at every geographic level.

I also contributed to the broader frontend module work, the API design, and the AWS architecture redesign. Plus mentorship of other engineers on Clean Architecture practices.

Platform-Wide Implementation

  • Serverless Core: AWS Lambda, API Gateway, DynamoDB single-table design, S3 pre-signed assets, SQS pipelines, Cognito JWT.
  • Data & ML: LightFM hybrid recommender, geospatial proximity weighting, social graph signals.
  • Frontend Experience: React 19 / Next.js with rich data visualizations, map search, responsive layouts.
  • Reliability: Containerized workers, background training jobs, structured logging and metrics.

Architecture Style

Clean Architecture throughout. Handlers, services, repositories. Business logic never touches HTTP or database code directly. Every function testable in isolation. The system can swap DynamoDB for Postgres without rewriting a single business rule.