Move the local Zillow data handling proesses to aForge cluster, and rewrite R scripts adopting Spark.
By creating a Spark "mirror" of the local codes, the whole data handling and analysis processes can be dramatically accelerated, thanks to Spark's optimization to large datasets.