Booz Allen Hamilton brings together FAST, SharePoint and Enterprise 2.0 platforms (he mentioned open source, but not sure what platform) to capture information about everything that their people were doing on projects and with their clients in order to provide a rich internal knowledgebase. Since they were using SharePoint document storage and the Enterprise 2.0 blogs and wikis for collaboration (plus some document storage), they created two silos of information: iShare (SharePoint) and hello.bah.com (E2.0). By adding FAST for federated searching across both repositories, the users don’t have to care where the information is, only that they can find it at the right time. This includes searching of employee profiles on hello.bah.com based on required skills, which will provide not just the employee profile, but links to the content that they’ve authored related to the skills search term. hello.bah.com also allows for the creation of communities, and content in SharePoint can be associated with a community to aggregate the information as well.
Like any services firm, access to information within Booz Allen Hamilton is critical, where that information includes documents, wiki pages, blog posts and relationships between people and that content. Enterprise search brings all that together.
Were they using FAST as a federated search or did they allow FAST to index both data stores? There are two differences in the approach, in a true federated approach you will find that there is no lag between data submission in the store and having it available in the federated search application. In a crawl and index solution, there is a definite lag.
I don’t recall him mentioning which way it was done.