Voice versus Text
The technology available to analyze written data has been around for a while now, underpinned by sophisticated search engines and phrase-recognition models like the ones used by news-reading machines. But according to Juan-Manuel Soto, CEO of Fonetic, a provider of speech and data linguistic analytics for trading floors, analyzing voice data presents significant challenges.
On the voice side, an average trader will produce around 2,000 hours of recorded calls per year, according to Fonetic’s data. If you multiply this figure by the number of traders a bank might have in a financial center—say, around 500—and then multiply that again by the number of financial centers a tier-one bank might have—say, five—it is feasible that such organizations generate in the region of 5 million to 6 million hours of calls every year.
For further information you can log into “Waters Technology” clicking here