AMI allows multi-criteria search requests (fields, documents, meta-data, dates, numbers) across Search Engines, Internet sites, Intranets, Databases, Content Management Systems and Newsgroups etc. Connection to data sources is performed by our generic connector and based on a protocol of questioning, which allows the user to define the methods of access to specific data sources. The setting up of questioning protocols can be easily implemented by the System Administrator.

AMI Software can access information in two ways, which are often complimentary to one another

1. Questioning; which allows the user to ask a question or formulate a query using varying degrees of language complexity
2. Navigation; which allows the user to go through a list, sometimes referred to as a tree, of documents. Navigation presents source locations using HTML and hypertext links.

AMI can use both methods.

With the first method, the technology behind the connectors makes it possible to understand and simulate the search behaviour of a user, even a user less experienced in creating search queries and scenarios, since all the functions of interrogation will always be fully utilised by AMI. In addition to this, directing several requests towards a particular source makes it possible to develop more complex queries automatically if required.

With the second method, the selected source will first be indexed by AMI, which will then be able to question its own index via a dedicated connector. A connector can be weighted to a higher or lower relevance level. Provided that the API of the source is serviceable, the connector will make full use of it. In other cases, it will generate an HTML form designed for human interrogation.

Access to protected sources (Passwords, Forms, etc), often referred to as the Invisible Web, can also be automated with the legitimate user access rights.