Various embodiments are generally directed to a dynamic, flexible acronym decoder. A user may submit a query via one of a plurality of user interfaces. Information describing the user may be received to generate a context vector for the user. The query may be processed against a database of terms using the context vector, a machine learning model, and content tags applied to terms in the database. Processing the queries against the database may return a result set, and the ML model may be used to compute a score for each result. The results may be ordered based on the scores and returned as responsive to the query.