Overwrite.AI Helps Real Estate Cos to Create Listings

After spending over two decades in real estate economics and investment, 44-year-old Ayman Alashkar set out on a journey to marry real estate and mathematics, which led him to found Overwrite.AI in 2019. The UAE-based startup aims to revolutionize the way real estate agents create property listings by giving them an intelligent, and efficient way to craft property descriptions.

“The idea was to converge mathematical precision with the art of property making,” explains Alashkar in an interview with Inc. Arabia. However, his entrepreneurial journey started in 2012.

Alashkar, whose interest in entrepreneurship began in 2012, wanted to delve deeper into the fast-changing technological landscape. In 2016, when AI was presented mostly in the form of Facebook chatbots, he saw an opportunity to integrate computer language to build Overwrite.AI. 

“While the conversation then wasn’t around generative AI, I saw it as a blend between behavioral sciences, mathematical precision, and the right art of programming. Since I had a solid foundation in mathematics, I recognized the potential this could open out to the sector and industry I was in,” Alashkar tells us.

He explains the idea was to automate the process of property marketing creation for real estate agents. In the fast-paced world of residential real estate, where homes are listed, sold, and replaced at lightning speed, real estate agents depend on the creation of appealing and informative listings to attract potential buyers. However, this process, repeated at high frequencies globally, was marred by inherent human errors.

The question was whether an automated system could be developed to assist real estate agents in crafting their property listings. The goal wasn't just efficiency but also monetization. This marked the inception of an ambitious journey towards rewriting the future of real estate marketing.

Pooling in his personal savings, Alashkar started working to build Overwrite. “We were building generative AI before the term gained its popularity in 2022,” he tells us.

But the journey wasn't without its challenges and setbacks. The first challenge was sourcing the data required to train the system. Instead of relying on potentially unreliable third-party data sources, they opted for a painstakingly constructed training corpus.

This corpus, carefully crafted, comprised a vast collection of grammatically structured data -- libraries and sub-libraries containing thousands upon thousands of entries.

The goal was to enable the system to seamlessly stitch together the requisite content, ensuring both grammatical and factual accuracy. This approach eliminated the looming risk of copyright liability and safeguarded against AI-generated hallucinations, as the entire system was under their control.

The second challenge was prompt engineering. Unlike tech giants with abundant resources, they couldn't afford the luxury of a freeform input field. They had to ensure that every interaction with the AI resulted in output that was accurate, relevant, and aligned with the user's objectives. This required meticulous planning, precision, and a resourceful approach that defied the odds.

The company has raised funds from a few angels who are in the real estate segment and space. “It made sense for us to raise funding from those people who are our customers and understand our problems,” explained Alashkar.

As the world of real estate continues to evolve, Overwrite stands as a bright example of how the fusion of mathematics, technology, and artistry can reshape an industry.