System and method for conducting transactions without human intervention using speech recognition technology

BACKGROUND OF THE INVENTION

[0001] 1. Field of the Invention

[0002] This invention relates generally to speech recognition technology and more particularly to a system and method for conducting transactions without human intervention using speech recognition technology to process customer transaction information.

[0003] 2. Description of the Background Art

[0004] Many businesses or service providers (hereinafter “service providers”) have implemented telephone-based systems that allow customers to call those service providers to place orders for goods or services or to conduct other types of transactions. One shortcoming of these telephone-based systems is that human operators typically answer incoming customer calls and process customer transactions. Not only are these human operators sometimes not very well trained, they also frequently place customers on hold, especially during peak hours, to complete transactions from prior calls. The result is that customers often become frustrated when trying to conduct transactions over the phone, so they hang up in the middle of their transactions, thus terminating those transactions and causing the service providers to lose that business.

[0005] VoiceXML (Registered Trademark, owned by IEEE Industry Standards and Technology Organization, filed Aug. 9, 2000) is a language for creating voice-user interfaces, particularly for telephone-based systems. For example, VoiceXML has been used to create VoiceXML application-based systems such as voice portals and voice service providers. These types of systems allow service providers to provide automated, telephone-based information retrieval services and other transaction-based services to customers where the customers do not have to interact with human operators.

[0006] One drawback to implementing a VoiceXML application-based system is that the service provider has to design and build the system essentially from scratch (or pay a third party to design and build the system). In most instances, this means that the service provider has to design and build the VoiceXML application, design and configure the server on which the application will run and integrate the server with the service provider's existing enterprise systems. Further, the service provider has to design and build a voice browser to enable customers to access the VoiceXML application server and conduct transactions remotely over an appropriate communications medium such as a public switched telephone network. These technical hurdles are time consuming and prohibitively expensive for many service providers.

SUMMARY OF THE INVENTION

[0007] One embodiment of a system for processing transaction instructions without human intervention includes a voice interpreter for receiving transaction information, in the form of voice utterances or DTMF commands, and for processing that transaction information, a business application server for receiving the processed transaction information and for generating transaction instructions, a connector manager for interfacing with an enterprise system and for transmitting the transaction instructions to the enterprise system and at least one housing designed to enclose the voice interpreter, the business application server and the connector manager. The embodiment also includes a telephony interface that allows a customer to access the system using any type of communications medium, including without limitation, a public switched telephone system, a private telephone network, a voice-over-IP packet network or any type of wireless network.

[0008] One advantage of this system is that it constitutes a “turn-key” automated transaction system. A service provider may implement the system by simply “plugging” the service provider's enterprise system(s) into the connector manager and the communications medium used to access the system into the telephony interface. By using this system, the service provider avoids having to design and build an automated transaction system from scratch, meaning that the service provider does not have to design and build a business application server that is integrated with the service provider's enterprise system(s) or design and build voice browsing functionality that enables customers to access the business application server and remotely conduct a transaction over an appropriate communications medium. The system therefore is a straightforward and cost-effective way for a service provider to implement an automated transaction system.

BRIEF DESCRIPTION OF THE DRAWINGS

P-0009[0009]FIG. 1 is a block diagram illustrating one embodiment of a system used to conduct a transaction without human intervention, according to the invention;

P-0010[0010]FIG. 2 is a block diagram illustrating one embodiment of the voice appliance of FIG. 1, according to the invention;

P-0011[0011]FIG. 3 is a block diagram illustrating one embodiment of the business application server of FIG. 1, according to the invention;

P-0012[0012]FIG. 4 is a block diagram illustrating one embodiment of the connector manager of FIG. 2, according to the invention; and

P-0013[0013]FIG. 5 shows a flow chart of method steps for conducting a transaction without human intervention, according to one embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

P-0014[0014]FIG. 1 is a block diagram illustrating one embodiment of a system <b>100</b> used to conduct a transaction without human intervention, according to the invention. Typical transactions may include, for example, purchasing a product or a service. As shown, system <b>100</b> may include, without limitation, a phone <b>110</b>, a public switched telephone network (PSTN) <b>120</b>, a voice appliance <b>140</b>, an analog phone switch <b>142</b>, a human operator <b>144</b>, local area network (LAN) <b>150</b> and an enterprise system <b>160</b>. Using phone <b>110</b>, a customer calls a service provider with whom the customer wants to conduct the transaction, and the call is routed through PSTN <b>120</b> to voice appliance <b>140</b>.

P-0015[0015] As described herein, once the customer is in communication with voice appliance <b>140</b>, the customer and voice appliance <b>140</b> participate in a “dialog,” during which the customer transmits all information relevant to the transaction (the “transaction information”) to voice appliance <b>140</b>. The transaction information may be in the form of voice utterances spoken into phone <b>110</b> and, optionally, dual-tone multi-frequency (DTMF) commands entered into phone <b>110</b>. As explained in further detail below in conjunction with FIG. 2, voice appliance <b>140</b> is configured to participate in the dialog with the customer, to process the transaction information provided by the customer, to generate transaction instructions based on the transaction information and to submit the transaction instructions to enterprise system <b>160</b>. Voice appliance <b>140</b> typically may reside on the premises of the service provider.

P-0016[0016] Voice appliance <b>140</b> is coupled to enterprise system <b>160</b> via an enterprise network, such as LAN <b>150</b>, which may be any type of packet-based network (e.g., TCP/IP, IPX/SPX or NetBEUI) over which data (e.g., the transaction instructions described herein) is transmitted between voice appliance <b>140</b> and enterprise system <b>160</b> using HTTP or other similar transport protocols. Alternatively, voice appliance <b>140</b> may be coupled directly to enterprise system <b>160</b> using any type of serial ports such as USB or RS-232 ports or parallel ports.

P-0017[0017] One feature of voice appliance <b>140</b> is that the customer can opt to by-pass the automated transaction process and to have his or her call routed directly to human operator <b>144</b> so that human operator <b>144</b> may process the customer's transaction. Under such circumstances, voice appliance <b>140</b> is configured to route the customer's call to human operator <b>144</b> via analog phone switch <b>142</b>, which is coupled to voice appliance <b>140</b>. Those skilled in the art will recognize that analog phone switch <b>142</b> may be any type of analog or digital device that couples voice appliance <b>140</b> to human operator <b>144</b>.

P-0018[0018] Enterprise system <b>160</b> is configured to receive the transaction instructions submitted by voice appliance <b>140</b> and to process those transaction instructions. Enterprise system <b>160</b> may be any type of transaction-based system used by the service provider. For example, if the service provider is a restaurant such as a pizza delivery restaurant, fast food restaurant or some type of dining-in restaurant, enterprise system <b>160</b> may be a point-of-sale system, a reservation system or customer relationship management (CRM) system. If the service provider is a financial institution, enterprise system <b>160</b> may be a CRM system or a financial/accounting system such as Oracle Financials or Siebel Finance. Those ordinarily skilled in the art will recognize that a given service provider may have more than one enterprise system <b>160</b> and that voice appliance <b>140</b> may be adapted to couple to multiple enterprise systems simultaneously.

P-0019[0019] Those ordinarily skilled in the art also will recognize that PSTN <b>120</b> may be any type of telephone network, including but not limited to, a private telephone network such as PBX, a voice-over-IP packet network, any type of wireless network or any other suitable communications medium. Further, phone <b>110</b> may be any type of telephony device that couples to the telephone network used in system <b>100</b>.

P-0020[0020] In alternative embodiments, an analog phone switch or any other similar analog or digital device may couple PSTN <b>120</b> to voice appliance <b>140</b>. In addition, phone <b>110</b> and PSTN <b>120</b> may be replaced with any type of non-telephony; microphone-based device that can be coupled to voice appliance <b>140</b> and configured to transmit voice utterances and, optionally, DTMF commands to voice appliance <b>140</b>. An example of such a microphone-based device is a speaker/microphone device of the sort typically found at fast-food restaurant drive-through.

P-0021[0021]FIG. 2 is a block diagram illustrating one embodiment of voice appliance <b>140</b> of FIG. 1, according to the invention. As shown, voice appliance <b>140</b> may include, without limitation, a housing <b>200</b>, a telephony interface <b>202</b>, a voice interpreter <b>204</b>, a text-to-speech (TTS) engine <b>206</b>, an audio engine <b>208</b>, a speech recognition (SR) engine <b>210</b>, a business application server <b>212</b> and a connector manager <b>214</b>. Housing <b>200</b> can be made of any type of suitable material such as plastic, metal or hard rubber. In one embodiment, housing <b>200</b> is sized to enclose telephone interface <b>202</b>, voice interpreter <b>204</b>, TTS engine <b>206</b>, audio engine <b>208</b>, SR engine <b>210</b>, business application server <b>212</b> and connector manager <b>214</b>. In alternative embodiments, two or more separate and/or related housings may enclose any number of these various components.

P-0022[0022] Telephony interface <b>202</b> integrates voice interpreter <b>204</b> with PSTN <b>120</b> of FIG. 1. More specifically, telephony interface <b>202</b> is configured to answer an incoming call from the customer, to initiate a session with voice interpreter <b>204</b> and to manage the communication protocols between PSTN <b>120</b> and voice appliance <b>140</b>. Further, telephony interface <b>202</b> is configured to receive requests for customer transaction information (in the form of audio output) from voice interpreter <b>204</b>, to transmit those requests to the customer via PSTN <b>120</b>, to receive customer transaction information (in the form of audio input and DTMF commands) from PSTN <b>120</b> and to transmit that information to voice interpreter <b>204</b> for processing. The functionality of telephony interface <b>202</b> may be implemented in hardware and/or software. Intel's Dialogic card is an example of a commonly used telephony interface product.

P-0023[0023] Voice interpreter <b>204</b> is configured to control the dialog between the customer and voice appliance <b>140</b> by processing voice-adapted programmable code (“voice script”) that resides in business application server <b>212</b>. The voice script may be based on any language used to create voice-user interfaces, such as VoiceXML. As explained in greater detail herein, the voice script sets forth the “flow” of the dialog between the customer and voice appliance <b>140</b>. The flow delineates the types of information needed from the customer to process the customer's transaction as well as the order in which that information should be solicited from the customer. More specifically, voice interpreter <b>204</b> is configured to request and receive the voice script from business application server <b>212</b>, to parse through and execute the instructions in the voice script, to generate requests for customer transaction information (in the form of audio output), to transmit those requests to telephony interface <b>202</b>, to process incoming customer transaction information (in the form of audio input or DTMF commands) received from telephony interface <b>202</b> in the form of audio input and to transmit the processed transaction information to business application server <b>212</b>. Voice interpreter <b>204</b> may be any VoiceXML interpreter or any other similar device.

P-0024[0024] When telephony interface <b>202</b> answers the incoming call from the customer and initiates a session with voice interpreter <b>204</b>, voice interpreter <b>204</b> requests the first portion of the voice script that resides in business application server <b>212</b>. Business application server <b>212</b> is configured to receive this request from voice interpreter <b>204</b> and to transmit the first portion of the voice script to voice interpreter <b>204</b> for processing. Voice interpreter <b>204</b> then parses through and executes the instructions in that first portion of voice script. For example, if the voice script indicates that voice appliance <b>140</b> should request certain transaction information from the customer, such as a selection from a group of choices or specific input relevant to the transaction at hand, voice interpreter <b>204</b> transmits that request to audio engine <b>208</b> for processing. Audio engine <b>208</b> may be any automated library of pre-recorded audio files and is configured to receive the transaction information request, to locate the pre-recorded audio file that matches the request and to transmit the contents of that audio file to voice interpreter <b>204</b>. In turn, voice interpreter <b>204</b> transmits as audio output the contents of the file to telephony interface <b>202</b> (where the contents are then transmitted or played to the customer via phone <b>110</b> and PSTN <b>120</b>). In the event that audio engine <b>208</b> cannot locate an audio file that matches the transaction information request, voice interpreter <b>204</b> may instead transmit the transaction information request to TTS engine <b>206</b> for processing. TTS engine <b>206</b> may be any standard speech synthesis engine and is configured to receive the transaction information request, to generate synthetic speech that matches the request and to transmit the synthetic speech to voice interpreter <b>204</b>. In turn, voice interpreter <b>204</b> transmits as audio output the synthetic speech to telephony interface <b>202</b> (where the synthetic speech is then transmitted or played to the customer via phone <b>110</b> and PSTN <b>120</b>).

P-0025[0025] Similarly, if the voice script indicates that the customer should transmit transaction information to voice appliance <b>140</b>, voice interpreter <b>204</b> directs the incoming transaction information that is in the form of audio input to SR engine <b>210</b> for processing. SR engine <b>210</b> may be any standard automated speech recognition engine and is configured to receive the audio input and to process the audio input by, among other things, interpreting the audio input and generating a data stream or equivalent set of information that matches the audio input. SR engine <b>210</b> is further configured to transmit the processed transaction information to voice interpreter <b>204</b>, which, in turn, transmits that information to business application server <b>212</b>. In the situation where the incoming transaction information is in the form of DTMF commands, voice interpreter <b>204</b> directs that transaction information to business application server <b>212</b> without first diverting the information to SR engine <b>210</b> for processing.

P-0026[0026] Voice interpreter <b>204</b> also is configured to analyze the flow set forth in the voice script and to determine whether additional dialog with the customer is necessary based on factors such as whether additional transaction information is needed from the customer to process the customer's transaction. If voice interpreter <b>204</b> determines that additional transaction information is needed, voice interpreter <b>204</b> requests from business application server <b>212</b> the next portion of the voice script as set forth in the flow. Business application server <b>212</b> is configured to receive this request from voice interpreter <b>204</b> and to transmit the next portion of the voice script to voice interpreter <b>204</b> for processing. Voice interpreter <b>204</b> receives this next portion of the voice script and parses through and executes the instructions contained in that portion of script. As previously described herein, the result of this process is that voice appliance <b>140</b> requests and receives additional transaction information from the customer. Again, voice interpreter <b>204</b> processes this transaction information and transmits it to business application server <b>212</b>. This process repeats until voice interpreter <b>204</b> determines that no further transaction information is needed from the customer to process the customer's transaction. All communications between voice interpreter <b>204</b> and business application server <b>212</b> take place using HTTP or other similar transport protocols.

P-0027[0027] As previously described herein, business application server <b>212</b> is configured to receive requests for portions of the voice script from voice interpreter <b>204</b>, to process those requests and transmit the requested portions of the voice script to voice interpreter <b>204</b> for processing and to receive the processed transaction information transmitted by voice interpreter <b>204</b>. Business application server <b>212</b> is further configured to compile this processed transaction information, to generate transaction instructions upon receiving all of the necessary transaction information from the customer and to transmit the transaction instructions to connector manager <b>214</b>. The transaction instructions may be implemented using XML or any other similar language or any type of object-based communications. As discussed in greater detail below in conjunction with FIG. 4, connector manager <b>214</b> is configured to receive the transaction instructions from business application server <b>212</b>, to translate those instructions into a format understood by enterprise system <b>160</b> and to transmit those instructions, via LAN <b>150</b> or directly, to enterprise system <b>160</b> for processing.

P-0028[0028] The form of the transaction instructions will vary according to the types of transactions that system <b>100</b> is designed to process. As those skilled in the art will recognize, the instructions contained in the voice script and the transaction-specific functionality of enterprise system <b>160</b> are two, but not necessarily the only, factors that define the form of the transaction instructions. For example, if the voice script sets forth a process for ordering a pizza, and enterprise system <b>160</b> is a point-of-sale system, then the transaction instructions may be an order for a particular type of pizza that the customer wants to eat for dinner. Similarly, if the voice script sets forth a process for setting up a 401(k) account, and enterprise system <b>160</b> is a system for storing and managing those accounts, then the transaction instructions may designate a new mutual fund that the customer wants to add to his or her 401(k) account or a new allocation of funds among the mutual funds in the customer's 401(k) account.

P-0029[0029]FIG. 3 is a block diagram illustrating one embodiment of business application server <b>212</b> of FIG. 1, according to the invention. As shown, business application server <b>212</b> may include, without limitation, a business application <b>300</b>, a remote administration module <b>306</b>, an appliance/module administration module <b>308</b> and a data store <b>310</b>. Business application server <b>212</b> may be any web server or similar computing device that is accessible using HTTP or any other similar protocols.

P-0030[0030] Among other things, business application <b>300</b> contains the voice script previously described herein. In one embodiment, business application <b>300</b> is an order-based application (i.e., a set of program instructions) that pizza delivery, take-out and dining-in restaurants, for example, may use. As also shown in FIG. 3, the order-based application includes, without limitation, takeout order module <b>302</b> and reservation module <b>304</b>. Take out order module <b>302</b> is configured to take a food order from a customer and, among other things, contains the portions of the voice script that set forth the flow for taking such food orders. The portions of the voice script contained in take out module <b>302</b> therefore delineate the types of information needed from the customer and the order in which that information should be solicited/requested from the customer to generate that customer's food order. For example, in the pizza delivery context, the voice script may set forth a series of questions asked to the customer to determine, among other things, the type of crust and the various toppings that the customer wants for his or her pizza. The voice script also may include questions pertaining to how the customer wants to pay for the pizza (e.g., credit card, debit card or cash) as well as delivery instructions and/or directions. In addition, the voice script may include instructions for transmitting certain information to the customer relevant to the customer's order, such as the cost of certain toppings or of different sizes of pizza, different order options that the customer may have as well as estimated delivery time.

P-0031[0031] Take out order module <b>302</b> may include various functionalities that enhance the overall effectiveness of the order-based application. For example, take out module <b>302</b> may include specific program instructions that provide for a caller identification functionality that identifies a repeat customer based on that customer's voice, phone number, DTMF commands or some other similar type of input. Take out module <b>302</b> also may include specific program instructions that provide for a repeat-order functionality that allows an identified repeat customer to circumvent the regular order-taking process and simply reorder one of the items ordered by that customer in one or more past transactions. Similarly, take out module <b>302</b> may include specific program instructions that provide for a functionality that confirms customer-based information such as delivery address and credit card information for identified repeat customers. Other functionalities that take out order module <b>302</b> may have include, without limitation, a suggestive selling functionality (where information regarding various types of promotions is communicated to customers), a special offer functionality (where customers are advised of additional items that they can purchase that will qualify those customers for various special offers or promotions) and a loyalty tracking functionality (where a point system or similar system is used to track customer order histories so that customers can qualify for special benefits).

P-0032[0032] Reservation module <b>304</b> is configured to take a reservation request from a customer and, among other things, contains the portions of the voice script that set forth the flow for taking such reservation requests. The portions of voice script contained in reservation module <b>304</b> therefore delineate the types of information needed from a customer and the order in which that information should be solicited/requested from the customer to generate that customer's reservation request. For example, in the dining-in restaurant context, the voice script may set forth a series of questions asked to the customer to determine, among other things, the time at which the customer would like to dine, the number of persons in the customer's party and the customer's table location preference. The voice script also may include informational transmissions to the customer that confirm the reservation time and the number of person in the customer's party.

P-0033[0033] Data store <b>310</b> is configured to store persistent data necessary to execute the voice script contained in business application <b>300</b>. Data store <b>310</b> may contain one or more databases, XML files or any other persistent data structures or storage mechanisms used to store data. For example, in the situation where business application <b>300</b> is an order-based application, data store <b>310</b> may contain, without limitation, the menus that a particular restaurant offers, the restaurant's pricing rules, information relating to the past orders of customers and statistics based on those past orders or past customers. Similarly, in the situation where business application <b>300</b> is a 401(k) account management application, data store <b>310</b> may contain, without limitation, listings of the various mutual funds in the 401(k) program, the fee structures of those mutual funds, information relating to past account choices made by program participants and statistics based on those past choices or past participants.

P-0034[0034] Those skilled in the art will recognize that in alternative embodiments business application <b>300</b> may be configured to access some or all of the data necessary to execute portions of the voice script from enterprise system <b>160</b> instead of or in addition to data store <b>310</b>. For example, in the situation where business application <b>300</b> is an order-based application and enterprise system <b>160</b> is a point-of-sales system, enterprise system <b>160</b> may store customer information such as credit card information, delivery address information or demographic information about the service provider's historic customer base. Enterprise system <b>160</b> also may store, without limitation, information relating to the past orders of customers, product information, the menus that a particular service provider offers as well as the pricing rules relating to the different products that the service provider offers.

P-0035[0035] Remote administration module <b>306</b> is configured to enable the remote administration of the different components of voice appliance <b>140</b> such as, for example, business application <b>300</b> and its relevant modules and connector manager <b>214</b>. Remote administration module <b>306</b> is further configured to manage connectivity to voice appliance <b>140</b> by a remote dial-in connection, by a scheduled, automatic dial-out connection or through a LAN-based connection. Once connected, a system administrator may service, manage or configure the different components of voice appliance <b>140</b> via remote administration module <b>306</b> using either terminal-based commands, a web-based interface such as a browser, or available software applications such as Microsoft's NetMeeting.

P-0036[0036]FIG. 4 is a block diagram illustrating one embodiment of connector manager <b>214</b> of FIG. 2, according to the invention. As shown, connector manager <b>214</b> may include, without limitation, one or more adaptors, such as adaptor <b>402</b>, adaptor <b>404</b> and adaptor <b>406</b>, enterprise system interface <b>408</b> and dial-up modem <b>410</b>. Generally, connector manager <b>214</b> is configured to translate information received from business application server <b>212</b> into a format that can be understood by enterprise system <b>160</b> and to translate information received from enterprise system <b>160</b> into a format that can by understood by business application server <b>212</b>. The translation functionality of connector manager <b>214</b> enables business application server <b>212</b> and enterprise system <b>160</b> to communicate with one another. More specifically, adaptors such as adaptor <b>402</b>, adaptor <b>404</b> and adaptor <b>406</b> provide connector manager <b>214</b> with this translation functionality. For example, each of adaptor <b>402</b>, adaptor <b>404</b> and adaptor <b>406</b> may be configured to interface with a unique type of commercial enterprise system such that each of adaptor <b>402</b>, adaptor <b>404</b> and adaptor <b>406</b>, as the case may be, is able to translate information received from business application server <b>212</b> into a format understood by a particular type of enterprise system as well as receive translate information received from that particular type of enterprise system into a format understood by business application server <b>212</b>. Examples of various types of adaptors include, but are not limited to, an adaptor configured to interface with a database enterprise system such as the Oracle 11i CRM system, an adaptor configured to interface with a point-of-sale enterprise system such as the Breakaway Relief Manager Plus system, an adaptor configured to interface with an enterprise system that supports EDI, an adaptor configured to interface with a printer and an adaptor configured to interface with a facsimile machine or any other similar type of device.

P-0037[0037] In one embodiment, the total number of adaptors <b>402</b>, <b>404</b> and <b>406</b> included in connector manager <b>214</b> is equal to the number of enterprise systems <b>160</b> in system <b>100</b> (i.e., system <b>100</b> has three enterprise systems <b>160</b>, each of which interfaces uniquely with one of adaptor <b>402</b>, adaptor <b>404</b> and adaptor <b>406</b>). Among other things, such an arrangement allows voice appliance <b>140</b> to be a “turn-key” device because the service provider can simply “plug” voice appliance into its existing enterprise system infrastructure by coupling each of adaptor <b>402</b>, adaptor <b>404</b> and adaptor <b>406</b> to the enterprise system <b>160</b> with which adaptor <b>402</b>, adaptor <b>404</b> or adaptor <b>406</b> has been uniquely configured to interface.

P-0038[0038] Connector manager <b>214</b> is further configured to manage the flow of information between business application server <b>212</b> and enterprise system <b>160</b> by (i) receiving information from business application server <b>212</b>, directing that information through the appropriate adaptor(s), such as adaptor <b>402</b>, adaptor <b>404</b> and/or adaptor <b>406</b>, and transmitting that information via enterprise system interface <b>408</b> to enterprise system <b>160</b> and (ii) receiving information from enterprise system <b>160</b> via enterprise system interface <b>408</b>, directing that information through the appropriate adaptor(s), such as adaptor <b>402</b>, adaptor <b>404</b> and/or adaptor <b>406</b>, and transmitting that information to business application server <b>212</b>. In addition, connector manager <b>214</b> is configured to manage the protocol(s) used to transmit information from enterprise system <b>160</b>. For example, connector manager <b>214</b> may transmit transaction instructions to enterprise system <b>160</b> using HTTP if those instructions are implemented using XML, or connector manager <b>214</b> may use SQL to transmit information to enterprise system <b>160</b> if enterprise system <b>160</b> is a database system. Other protocols that connector manager <b>214</b> may use include TCP/IP or any other suitable protocol or language. The functionality of connector manager <b>214</b> and adaptor <b>402</b>, adaptor <b>404</b> and adaptor <b>406</b> (as well as any other adaptors) may be implemented in hardware and/or software.

P-0039[0039] Enterprise system interface <b>408</b> is configured to couple connector manager <b>214</b> to LAN <b>150</b>, where voice appliance <b>140</b> is coupled to enterprise system <b>160</b> indirectly via LAN <b>150</b>, or to couple connector manager <b>214</b> to enterprise system <b>160</b>, where voice appliance <b>140</b> is coupled to enterprise system <b>160</b> directly. In the former situation, enterprise system interface <b>408</b> may be any type of appropriate network interface card such as an OC-3 SONET connection or an Ethernet over fiber connection. In the latter situation, enterprise interface <b>408</b> may be any type of serial port such as a USB or RS-232 port or any type of parallel port.

P-0040[0040] Dial-up modem <b>410</b> is the device through which remote dial-in connections and automatic, dial-out connections occur for purposes of remotely administering voice appliance <b>140</b> as previously described herein. Dial-up modem <b>410</b> may be any type of modem or similar communication device. Those skilled in the art will recognize that in alternative embodiments, dial-up modem <b>410</b> may reside outside of connector manager <b>214</b> and be located anywhere within or external to voice appliance <b>140</b>. Further, dial-up modem <b>410</b> can be substituted with any other suitable communications interface known in the art to effectuate remote administration.

P-0041[0041]FIG. 5 shows a flowchart of method steps for conducting a transaction without human intervention, according to one embodiment of the invention. Although the method steps are described in the context of the systems illustrated in FIGS. <b>1</b>-<b>4</b>, any system configured to perform the methods steps is within the scope of the invention.

P-0042[0042] As shown in FIG. 5, the method for conducting a transaction without human intervention starts in step <b>510</b> where voice appliance <b>140</b> requests transaction information from a customer. As described herein, in one embodiment, the customer accesses voice appliance <b>140</b> by calling via phone <b>110</b> the service provider with whom the customer wants to conduct the transaction. Once in communication with voice appliance <b>140</b>, voice interpreter <b>204</b> requests from business application server <b>212</b> the first portion of the voice script contained in business application <b>300</b>, which resides in business application server <b>212</b>. Voice interpreter <b>204</b> parses through and executes the instructions in this first portion of voice script. These instructions include requesting certain transaction information from the customer. The requests for transaction information are played/transmitted from voice interpreter <b>204</b> to the customer using audio engine <b>208</b> and/or TTS engine <b>206</b>.

P-0043[0043] In step <b>512</b>, voice appliance <b>140</b> receives the transaction information requested from the customer. The transaction information may be in the form of voice utterances spoken into phone <b>110</b> and, optionally, DTMF commands entered into phone <b>110</b>. In step <b>514</b>, voice interpreter <b>204</b> processes the received transaction information using SR engine <b>210</b>, to the extent that the transaction information is in the form of voice utterances, and transmits the processed transaction information to business application server <b>212</b>. In step <b>516</b>, voice interpreter <b>204</b> analyzes the flow set forth in the voice script and determines whether any addition transaction information is needed from the customer to process the customer's transaction.

P-0044[0044] If voice interpreter <b>204</b> determines that additional transaction information is needed from the customer, voice interpreter <b>204</b> requests the next portion of the voice script, which contains instructions for requesting additional transaction information from the customer, from business application server <b>212</b> and the method returns to step <b>510</b>. If voice interpreter <b>204</b> determines that no further transaction information is needed from the customer, then in step <b>518</b>, business application server <b>212</b> compiles the processed transaction information received from voice interpreter <b>204</b> and generates transaction instructions. In step <b>520</b>, business application server <b>212</b> via connector manager <b>214</b> transmits or submits the transaction instructions to enterprise system <b>160</b> for processing. In step <b>522</b>, enterprise system <b>160</b> processes the transaction instructions.

P-0045[0045] One advantage of the system (and associated methods) described above is that it constitutes a “turn-key” automated transaction system. A service provider may implement the functionality of voice appliance <b>140</b> by simply “plugging” the service provider's enterprise system(s) <b>160</b> into connector manager <b>214</b> and the communications medium used to access voice appliance <b>140</b> into telephony interface <b>202</b>. By using voice appliance <b>140</b>, the service provider avoids having to design and build an automated transaction system from scratch, meaning that the service provider does not have to design and build business application server <b>212</b> that is integrated with the service provider's enterprise system(s) <b>160</b> or design and build voice browsing functionality that enables customers to access business application server <b>212</b> and remotely conduct a transaction over an appropriate communications medium. The system therefore is a straightforward and cost-effective way for a service provider to implement an automated transaction system.

P-0046[0046] The invention has been described above with reference to specific embodiments. One skilled in the art will recognize, however, that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention as set forth in the appended claims. For example, telephony interface <b>202</b>, voice interpreter <b>204</b> (as well as TTS engine <b>206</b>, audio engine <b>208</b> and SR engine <b>210</b>), business application server <b>212</b> and connector manager <b>214</b> may run on a common processor or hardware platform. Alternatively, voice appliance <b>140</b> may be designed such that one or more of these components may run on one or more separate processors or hardware platforms. Also, one or more business applications <b>300</b> may reside in business application server <b>212</b>. This capability allows a service provider to use one voice appliance <b>140</b> to conduct different types of transactions simultaneously or in series without having to introduce additional business applications servers <b>212</b> into voice appliance <b>140</b> or having to use more than one voice appliance <b>140</b> in system <b>100</b>. In addition, voice appliance <b>140</b> may be implemented using a distributed architecture. For example, suppose a service provider has three locations at which the service provider wants to set up automated transactions systems <b>100</b>. One could design voice appliance <b>140</b> such that a separate set of telephony interface <b>202</b> and voice interpreter <b>202</b> (along with TTS engine <b>206</b>, audio engine <b>208</b> and SR engine <b>210</b>) resides at each of the three locations, and each set of telephony interface <b>202</b> and voice interpreter <b>204</b> communicates to one centrally located business application server <b>212</b> and connector manager <b>214</b>. The foregoing description and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.