GeoMetro Invest

GeoMetro Invest is a real estate intelligence platform that scans listings across Brazil, benchmarks prices per m², and uncovers undervalued assets—turning data-driven insights into superior ROI.

Comment

About the Project

This project was born at the intersection of timing, curiosity, and opportunity.

I am the CEO of STOIX.dev, a software engineer with 15 years of experience in technology, and a long-time programming enthusiast. In recent years, I have dedicated my free time to experimenting with AI systems and learning how to design robust and resilient architectures in a practical, fast, and pragmatic way.

During a conversation with a commercial partner, I was introduced to a real estate business model that immediately caught my attention. It was an opportunity to participate in a startup that had already been operating actively in the Brazilian real estate market, applying a proven investment methodology. This idea was originally on my roadmap for 2026, but with the momentum of HackTown and the quiet days at the end of the year, I decided to accelerate the plan and bring it forward.

Inspiration

What inspired me most was realizing how much AI amplifies the leverage of a high-level engineer today. The ability to go from idea to production-grade systems in days instead of months is transformative. Beyond the technical excitement, there was also a strategic incentive: by contributing technology to automate and scale an already profitable operation, I likely secured my stake in a brokerage company that has been growing steadily for the past four years.

How I Built It

The system was developed using a microservices-first approach, focused on speed, scalability, and resilience:

  • Warp was used to rapidly design and deploy crawler microservices responsible for collecting and normalizing real estate listing data.
  • Lovable was used to build the frontend quickly, allowing fast iteration and validation.
  • AI-assisted development (vibe coding) played a central role in accelerating architecture decisions, implementation, and refactoring.

The result is an automated pipeline that systematizes a real-world investment strategy already applied to more than BRL 50 million in transactions.

What I Learned

  • AI is not just a productivity tool—it is a force multiplier for system design and execution.
  • Clear domain understanding combined with automation can unlock massive operational scale.
  • Building fast does not mean building fragile, if architecture decisions are intentional.

Challenges

  • Translating a human-driven investment intuition into deterministic, automated systems.
  • Ensuring data quality and consistency across heterogeneous real estate sources.
  • Balancing speed of development with long-term maintainability.

What’s Next

The next step is to build a CRM platform for real estate agents, inspired by solutions like Zillow and Brivity (which raised over \$100k from Goldman Sachs). The goal is to address structural gaps in the Brazilian brokerage market.

This CRM will integrate with messaging platforms I am already a partner in:

  • smsfunnel.com.br
  • call4u.com.br

These integrations will maximize showing attendance, automate agent workflows, and increase conversion rates.

Looking further ahead, once a strong base of agents has been using the platform for approximately five years, the vision expands into residential services orchestration. In this model, real estate agents act as regional supervisors of service providers, earning commissions as area operators—similar to how Ecotrak operates in industrial maintenance.

This project represents more than software: it is a long-term bet on data, automation, and AI-driven leverage in the Brazilian real estate ecosystem.

Built With

Share this project:

Updates