This happens frequently in oral expressions. Yes, this is a "contraction" as you stated. The context is specific to everybody, so the speaker can shorten some words to express more information. Here the full word /xiànzài/ is shortened as something like /xià(n)ài/, where the sound /z/ and the nasal sound are in fact pronounced but in a very short period, and the recording seems like that the one just omitted /z/, /n/. You could try yourself by reading the word "现在" as quick as possible without altering the leading consonants [or, to be more exact, 声母] the tone and the vowels [韵母], and you might get the same result.