Many companies are now offering voice APIs for their telephony infrastructure, making it possible to create so-called voice mashups. One of the first was Jaduka, another that has gotten a lot of press is Ribbit.

APIs are necessary to create voice mashups. They allow you to treat voice and data interchangeably, building applications like an erector set. Twitterfone and Fonolo are great examples of voice mashups.

Is there money in this? I don't believe so. Read the piece I wrote for GigaOM: Is There Money in Voice APIs?