SpeechSwitch, Inc. was incorporated in New Jersey on November 10, 2004 as a wholly-owned subsidiary of iVoice, Inc. It is engaged in the design, manufacture, and marketing of specialized telecommunication equipment. Our products use standard open-architecture PC platforms and Microsoft Windows 2000 operating systems, thereby facilitating the rapid adoption of new PC-based technologies while reducing overall product costs. We concentrate our product development efforts on software rather than hardware because we believe that the most efficient way to create product value is to emphasize software solutions that meet customers' needs.

We have recently adapted our applications to integrate with different manufacturer telephone switches through the use of Telephony Application Program Interface or "TAPI". The use of TAPI, allows SpeechSwitch to integrate our applications into different telephone manufacturers Private Branch Exchange systems or “PBX's”, eliminating the need for costly additional external hardware. We have traditionally used standard PC-related hardware components in our products, in part, to limit our need to manufacture components. Our manufacturing operations consist only of the installation of our proprietary software and, if required, a voiceboard, into a fully assembled PC system which we obtain from several different vendors. The Company obtains system components such as PCs, circuit boards, application cards, faxboards, and voiceboards from various suppliers.

Our flagship product is our Speech-enabled Auto Attendant product. The Auto Attendant engages callers in a natural language dialog and is ready to transfer a caller to an extension for the party the caller is trying to reach at any time. Callers can interrupt the Auto Attendant at any time by barging in on the prompts and simply saying the name of the person or department they wish to speak to.

We have met interoperability standards with several leading PBX manufacturers for Auto Attendant. To date, rigorous testing and compatibility studies have developed into co-marketing arrangements with 3Com, for its NBX® platform, , and AltiGen’s AltiServ® for its phone systems. These recent platform integrations add to several others previously completed, including a Siemens Ready™ certification, NEC Fusion Strategic Alliance and Sprint North Supply. Through these co-marketing arrangements and strategic alliances, we will attempt to capture significant market share in the business communication solution market by expanding distribution through these manufacturers’ authorized reseller networks. To date, no significant revenues have been generated as a result of any such co-marketing arrangements or alliances.

Our Name Dialer product is an automatic phone dialing system. The system imports the necessary contact information for dialing (names and phone numbers) from a variety of sources including, but not limited to, Microsoft Outlook, ACT, and Gold Mine. The imported names are then transcribed, through software, into a set of phonemes that are used for voice recognition. When the end user picks up the handset, the call is automatically transferred through the PBX, to the Name Dialer application running on a server machine. The user simply says the name of the person (whose name came from the contact list) and the Name Dialer places the call.

Our Speech SDK product is a unique tool for software application developers which provides the ability to convert common command and control functions to speech commands. The SDK allows software developers to write applications that can treat a user's voice as an input device, such as mouse, keyboard, or joystick. In addition to telephony applications, the SDK can be used to incorporate speech recognition into games, handheld devices and even household appliances.

SpeechSwitch will market its products directly, with a sales force, and through more than 100 domestic and international re-sellers. SpeechSwitch intends to enter into arrangements with resellers to broaden distribution channels and to increase its sales penetration to specific markets and industries. Distributors will be selected based on their access to the markets, industries and customers that are candidates for the products.

The two main functionalities that make us unique are Adaptation and Tuning:

Adaptation

Adaptation, developed by SpeechSwitch, is a speech recognition system that continuously analyzes incoming speech input transforming the existing speech model to customized speaker models. The models produced by the technique, compensate for speaker accents, telephone line quality, and background noise at the customer site thereby boosting recognition accuracy. As the model aligns itself to the customer environment, speed of recognition per channel is greatly increased. The entire process is automated and the models generated load themselves into the system without any user interaction. The adaptation process can be performed any number of times to re-train the system if the above conditions change. Over a period of time, the trained model results in much faster and exceptional speech recognition accuracy at the customer end.

Tuning

The Channel tuning tool is an iterative data analysis and performance improvement tool for gathering and analyzing speech input over phone lines. The tool is designed to eliminate the time consuming manual process of examining the quality and volume of sound at the client site, accordingly calculate, adjust settings on both the sound processing hardware and the speech engine. The tool matches pre-calculated set of data obtained from recording files in ideal conditions with data calculated by re-playing the same set of files over the phone lines on the client side. The new data is used as a starting point to estimate quality of sound input to the speech engine. Several passes are made before analyzing the data gathered. The data is then matched with appropriate hardware settings and the new settings are then calculated. These settings are used as a starting point for the next set of iterations. The channel tuning tool is designed as a client-server component.

The compatibility with many phone systems and our dedication to our customers to meet and exceed customer expectation, the ability to develop to the client’s need and our flexibility in the development cycle, make clients feel at ease.

These are just a few of the systems we are compatible with:

SpeechSwitch Management