Cura | Devpost

Inspiration

Only 21% of real estate agents in the U.S. work with commercial properties, leaving most people without the knowledge or access they need. We built our modern property management system to change that.

Everyone should be able to explore, understand, and manage real estate with confidence, whether you’re an experienced investor or just starting out. Our platform gives people the clarity and control they deserve, helping them make smarter decisions and uncover real opportunities.

Real estate should be open, simple, and accessible to everyone. That’s what we’re building.

What it does

Cura provides forecasting and analytical insights, commercial real estate properties, pros and cons analysis, and market sentiment. It helps users make informed investment decisions through data-driven insights, interactive property comparisons, and real-time market trends. With Cura, users can explore opportunities, evaluate risks, and understand how each property aligns with their financial goals. Whether managing a portfolio or searching for a first commercial space, Cura brings clarity, confidence, and strategy to every decision.

How we built it

Cura is built around agentic-driven navigation. The system is designed to think and reason through user input, adapting layouts and responses based on context. From reasoning to web search to analytics, Cura connects each step to create a seamless and intelligent user experience.

Challenges we ran into

Integrating multiple data sources reliably Handling dynamic web scraping and structured data storage Maintaining real-time performance while using reasoning agents Building a smooth interaction flow between the AI and UI

Accomplishments that we're proud of

Developed a reasoning-based layout engine for property insights Built an analytical dashboard that visualizes property data in real time Integrated market sentiment and forecasting into one cohesive system

What we learned

The value of version control through consistent git commits and branching How to create and deploy Nemo models Data web scraping techniques and best practices Building with Crew AI, TypeScript, Python, FastAPI, MongoDB MCP, and Next.js

What’s next for Cura

Integrate a voice-based agent to handle property negotiations and client communication Expand property coverage and improve data freshness Scale infrastructure for larger datasets and real-time analytics