Enterprise Speech Reco-gnition (ESR) has set the metre running after scoring a lucrative four-year contract with Sydney's largest taxi company, Combined Communications Network (CCN).
Under the terms of the agreement, the Sydney-based voice recognition and integration company will build CCN an automated phone taxi ordering system, which ESR will provide via an application service provider (ASP) model priced on a per-call basis.
ESR officials claim the program will field around 60 per cent of Sydney's taxi phone orders, with the system designed to function as a live operator speaking in an Australian accent. The system will then instantly confirm the customer's order by SMS, fax or e-mail.
The company will integrate technology from voice recognition software developers Nuance Communications and Edify, as well as leverage technology from US-based voice recognition ASP Voci Corporation. ESR and Voci have entered into a strategic alliance to extend Voci's presence into the Asia-Pacific market, while providing ESR with technology, experience and a ready customer base, ESR said in a prepared statement.
According to Maria Leftakis, managing director of ESR, the company's engineers had a "good time" building the system because of the number of variations and the colloquial phrasing included in the program, which ESR will now brand as its Recogniser platform.
Leftakis claims the system has been programmed with 29,393 Sydney street names each pronounced four different ways, and 780 suburb names each recorded with three different intonations. It contains 81 responses for "yes" including such classics as "I reckon," "bloody oath" and "ripper", and 31 responses for "no".
But the burning question on many Sydneysider's minds is whether the solution will understand the drunken slur of a potential customer. Officials from both Nuance and Edify told ARN the system designed by ESR would "more than likely" have a good chance of understanding drunken idiot speech: "It works. You can't trick it, it's that good," said one official.
ESR also heralded its official opening last week with the appointment of former NSW premier and investment banker Neville Wran as the company's chairman.
Wran opened proceedings for ESR by expressing the feelings of many about getting lost inside a maze-like push button system or Interactive Voice Response (IVR).
"I think it's an understatement to say we will all be glad to see the last of those IVR telephone menus," Wran said.
Wran waxed lyrical about the nature of technology emulating its creators before endorsing what he referred to as "Australia's first - and only - speech recognition ASP".
ESR signed the contract in August and Leftakis claims the system, which is currently in a beta testing stage, will go live in around six weeks, handling up to 20 million calls annually.