Data records in one or more databases may be scanned to create a search index. The search index may include an identifier for each data record scanned, and a list of fragments of data in one or more data fields of each scanned data record. Each fragment in the list may be created by extracting a fixed number of adjoining characters from data in each data field of a corresponding data record. The list of fragments may include all fixed number of adjoining character sequences in each data field of a data record. Multiple search indexes may be included, with each search index including a different fixed number of adjoining characters in the fragment list. Systems and methods may be provided.