A group of PhDs and PostDocs at Harvard SEAS and IACS was working on real-estate predictions for 2 years time. We used a diverse set of inputs to accurately predict home values and the listings timeline. We provided a very sophisticated machine learning models that use publicly available data to predict prices, days-on-market, and leads. Instead of targeting all at once, brokers and realtors can use their marketing strategy more effectively. Buyers and sellers are able to get the exact properties estimates for their clients.