Ever wondered how an AI application transitions from a simple idea to a fully functional product? George Kamtziridis, Tech Lead at fromScratch Studio, took attendees through the complete journey, from identifying a problem, conducting research, and building datasets, to developing machine learning models and scaling them into a practical digital platform.
We spotlighted ReValue, a real-world project on noise pollution and house price estimation, where we created an entirely new open-source dataset for noise pollution, investigated its impact on house prices with machine learning (leading to a research paper), and turned our findings into a digital platform for estimating housing prices with noise pollution as a key factor. Attendees followed this end-to-end process and got access to the dataset and methodology via a GitHub repository.