Singer Equities is a private equity firm that owns and operates about a dozen distributors of industrial supply companies around the United States. I have built several eCommerce and brochure sites for two of their locations since 2008.
Each of the companies within Singer Equities was once an independent operation with its own inventory system. These systems do not communicate with each other and are not easily accessed by employees at different locations. Because of this, companies were double-ordering inventory already in stock at other locations.
SearchSinger.net aggregates a daily inventory report from each location which is then indexed by an Apache Solr instance.
Solr is a very fast and powerful open source search engine. It's great at handling massive data sets and can filter results in very specific ways. In this application, we index a MySQL database and we filter results by city, manufacturer, part number and availability.
Solr's real power is its ability to parse very long part numbers. Slight variations in these numbers could make it difficult to find items from different locations without an exact match. Solr includes a powerful auto-suggest feature that removes this barrier and makes finding things very easy, even if there are typos or inconsistencies in naming.
With Solr and various support programs, two hundred employees from ten locations can search 130,000+ items via a web interface, greatly reducing the instances of duplicate inventory being ordered.