Join us for overview of the differences between a traditional software development team and a speech application development team. Possible discussion topics will include:

• How to staff a speech application development team (e.g., hire consultants temporarily, hire new employees, or retrain existing employees)
• How to improve communication among team members
• What to do when projects are behind schedule

A customer concerned that their application was not performing drove us to explore new ways of analyzing system performance and conveying the results to our customers

Speech analytics is moving from a tool for governmental intelligence work to a commercially viable proposition in customer service. Join this session to learn of speech analytics and the implications of the technology.

As more businesses and consumers depend on speech technology, it’s becoming increasingly important to protect your investments. With more than fifteen years of experience in exploring computer security, Kevin Mitnick is a largely self-taught expert in exposing the vulnerabilities of complex operating systems and telecommunications devices. As the world’s most famous (former) hacker, Kevin has been the subject of countless news and magazine articles published throughout the world. Kevin will share how hackers compromise your security and what you can do about it.

Today’s consumers are increasingly sophisticated in their expectations for customer service from the organizations with which they do business. They’re also increasingly willing to publicize negative experiences. This combination leads to a volatile marketplace in which organizations must protect their customers against poor user experiences to establish and maintain their competitive advantage. Experts in this session discuss how to realistically assess customer satisfaction and deliver a positive experience to every customer, every time.

Organizations can no longer hide from their customers’ bad experiences. Expectations are higher than ever before, and with today’s social media, the consumer’s voice is louder and travels further. That’s why you must ensure customers are satisfied with their increasingly automated and mobile interactions with your company. This presentation will talk about the shift of power from company to consumer. It’s also designed to show the impact social media is having on customer care and the opportunity businesses have to make favorable impressions by leveraging the power of speech.

The business success of automated voice services depends significantly on a well-founded understanding of how they are used by customers. Modern tools for dialogue analysis allow business units to produce meaningful reports that provide a 360-degree view of customer interactions. These reports help call centers control and optimize their applications and systems and are an essential part of achieving broad corporate objectives, including greater customer satisfaction, cost reductions, and increased revenue. The session will identify caller experience information that organizations must take into account.

How should an organization decide between speech and touchtone IVR solutions? The thinking used to be that speech was an all-or-nothing decision, but increasingly organizations are discovering that a speech-plus-touchtone solution may be the best fit for their IVRs. Learn how to decide which features require speech, which may be well-served by touchtone, and how to create a smooth, cohesive user interface that includes both modalities.

Buyers, beware: yYu might not need a big, expensive speech application when a simple touch-tone application could be the perfect fit. Or perhaps you do need the new speech application, but with a twist. No matter, the perfect always requires a solid case and strategy for what’s purchased and deployed. We’ll explore what you must consider and how to determine what best suits the needs of your business and your callers.

DTMF applications are usually easier, faster, and cheaper to develop than speech applications. As waves of proprietary IVR systems and hardware platforms reach their end-of-life, they will need to be replaced. As a result, many businesses turn to vendors to replace those systems in a like-for-like fashion—an approach that might saves time and money, but does it enhance customer loyalty and represent a service that meets callers’ needs? This presentation will explore a few DTMF-to-DTMF application rewrites and how the same business needs could have been better addressed with speech.

How secure is your speech application? By now, organizations are accustomed to dealing with security threats that arrive via email and Web channels, but they are sometimes less familiar with the threats presented by the use of speech technologies. In this session you will learn how to identify emerging threats to the speech channel and to develop strategies that effectively protect your data, your customers, and your organization.

This session will cover current capabilities and market trends related to improving security for self-service environments. We will discuss such topics as federal regulations that are driving deployments (such as multifactor authentication), current technology capabilities, and industry trends. Case studies will highlight uses of security technologies and considerations for production deployments.

Identity theft costs us an estimate $53 billion annually, and VoIP phishing is simply the latest assault by hackers to steal personal information. It exploits the trust people have in IVRs, toll-free numbers, and strong brands. In this presentation, we will witness a live demonstration of VoIP phishing, get under the hood to see exactly how hackers accomplish it, and explain what companies should do to reduce the risk to their customers and, ultimately, their bottom lines.

Experts in this panel will gaze into the crystal ball and offer their predictions for new developments in speech technologies. Join us for a spirited discussion of upcoming trends in speech recognition, speaker biometrics, and speech analytics. This session will prepare you for what’s next in the world of speech and explain how to be one step ahead of the curve in making the most of speech technologies.

As more businesses and consumers depend on speech technology, it’s becoming increasingly important to protect your investments. With more than fifteen years of experience in exploring computer security, Kevin Mitnick is a largely self-taught expert in exposing the vulnerabilities of complex operating systems and telecommunications devices. As the world’s most famous (former) hacker, Kevin has been the subject of countless news and magazine articles published throughout the world. Kevin will share how hackers compromise your security and what you can do about it.

This session brings together two different topics in VUI design. In the first presentation, we explore yes/no questions—an extremely common but troublesome element in VUI design. The second presentation switches gears to discuss the ways in which design must change to meet the needs of private VUIs, meant for users within an organization, versus public VUIs. These presentations will ask specific questions and give targeted answers that you can apply to your VUI.

A large variety of yes/no questions are fundamental in today’s speech applications. Callers offer widely differing responses, which can be a challenge to commonly-used, generic yes/no grammars. In this session you will learn about research into the types of caller responses, implications for dialogue design, and how to build robust templates that can handle the variation.

Are there a different set of rules that should be followed for a private-facing application versus a public-facing one? This discussion will focus on lessons learned in designing a personalized, private-facing application, detailing where its design overlaps with a public-facing app and where the needs of private users are clearly different. You will also hear how presence and multimodality may affect where private-facing applications could be headed.

In the United States, the Spanish language portion of a speech project is often an afterthought. Organizations recognize the need to offer telephony services in Spanish, but they do not always devote the resources and attention necessary to ensure a positive and productive experience. The experts in this session will present proactive strategies that ensure your Spanish voice interaction meets the unique needs of Hispanic customers and offers them a compelling reason to use automation.

A caller’s initial interaction with a multilingual application is in choosing the language for communicating with the system, typically English or Spanish. The presentation of language choice can affect system performance, customer satisfaction, and company image. Diverse issues, such as the political climate, immigration controversies, and the quality of Spanish applications, affect companies’ willingness to offer Spanish, as well as callers’ willingness to use it. This session presents data from existing applications that illustrate the need for improvement.

Many companies offer customers the choice of using Spanish, but do not understand the issue of "Spanglish." Spanish speakers in the United States often mix English and Spanish words in the same sentence, or mix languages to create new words. This has deep implications for the design of grammars and user dictionaries, as well as the use of parallel grammars. This session will propose methods for understanding and addressing these issues to create a successful Spanish system.

Can one user interface encompass both speech and touchtone input in an efficient, effective, and satisfying interaction? This session explores this question and offers insight into the ways speech and touchtone can work in concert. Learn techniques for writing prompts and call flows that make sense across modalities, and understand the data that supports these techniques. This session will teach you how to leverage both speech and touchtone so yor organization can meet its business objectives.

This session presents research on user interfaces that combine the stability of touchtone with the elegance of speech recognition. With results from real-world testing, you'll learn how modality preference varies among callers, by context for an individual caller, and even on a turn-by-turn basis in the dialogue. We present a behaviorist approach to show how interfaces can rapidly adapt and with no overhead in terms of task completion time.

This presentation reviews the pros and cons of designing and deploying a "press or say" speech application. While some believe this is an easy way to transition from DTMF to speech applications in the enterprise, others consider it a bad idea for a variety of reasons. This session will detail recent findings to help you make the best choice.

What are the hot topics in VUI today? Where do we agree, and where do controversies still exist? Join us for the third annual panel debate on the state of the art in VUI design, refereed again by Syntellect’s Greg Simsar. Come with questions for this year’s panel of VUI practitioners and be ready to share your two cents in spirited discussions about where VUI stands as a profession and a discipline.

As more businesses and consumers depend on speech technology, it’s becoming increasingly important to protect your investments. With more than fifteen years of experience in exploring computer security, Kevin Mitnick is a largely self-taught expert in exposing the vulnerabilities of complex operating systems and telecommunications devices. As the world’s most famous (former) hacker, Kevin has been the subject of countless news and magazine articles published throughout the world. Kevin will share how hackers compromise your security and what you can do about it.

Speech-enabled applications become more effective by displaying a visual display or video at the same time that the user speaks and listens. The phone display may present menus, pictures of products, live video clips, and video commercials. This session discusses how to harness video in VXML to create rich voice and video applications, as well as how to author multimodal applications for these devices.

With the increasing proliferation of fixed and mobile videophones, and media-friendly next-generation networks such as IMS coming closer to a reality, the market waits for 1-D audio services to blossom into interactive, personable, engaging multimedia interfaces. In this presentation we show how to effectively harness video-in-VXML to quickly create rich voice and video applications and discuss some considerations when implementing a wide range of new services such as video mail, adult services, video dating, TV-on-mobile, video contact centers, etc.

Around the world consumers have access to live two-way video using 3G cell phones, broadband/DSL videophones, and video telephone software on their personal computers. This talk will show how speech-enabled self-service applications can become even more effective by displaying to a caller a visual display or video at the same time that they use the speech application menus, pictures of products, live video clips, and video commercials. Specific application development techniques that leverage speech and visuals together will be presented, along with sample speech-enabled voice-plus-video applications for universities, travel, retail, and home health care. The process of application authoring will be part of the talk.

Speech technology is increasingly being used in the automobile. The first case study in this session will show how a speech-enabled, in-vehicle device can read email and text messages aloud and enable users to instantly respond by voice. This session’s second case study will reveal how customers can use their mobile phones to retrieve location and current real-time parking availability at parking facilities; make parking reservations at a facility; pay for parking; and provision a new user account and provide outbound informational alerts.

A new, state-of-the-art, speech-enabled in-vehicle device exists that not only reads email and text messages aloud, but enables you to instantly respond to them by voice. Generated speech notifies drivers with timely information about phone calls, calendar appointments, and other potentially distracting mobile phone events. The following challenges will be covered: in-vehicle voice interfaces require extremely careful design and focus around the user's primary task—driving, and flexibility, personalization, and appropriateness of dialogue content are important in a VUI used on a daily basis.

ParkingCarma needed a scalable and safe way to enable its customers to access its services while driving. Companies should be able to perform the following functions in as cost-effective and convenient manner possible through a speech enabled application over a mobile phone: search for and get the location and current real-time parking availability at parking facilities, make a reservation for parking at a facility, pay for parking at a facility, and provision a new user account and provide outbound informational alerts. The system also had to be designed to take into account the harsh noise environment inside of cars and parking facilities.

A multimodal application is more than just a multimedia application. A multimodal application supports multiple techniques for entering information. This presentation will showcase standards-based multimodal applications developed by W3C Multimodal Interaction Working Group member companies. They will also demonstrate some laboratory applications that illustrate the principles of the W3C Multimodal Architecture and a deployed healthcare application that enables users to type, speak, and scribble on their mobile devices to record vital readings, highlight regions of interest on images, and submit them wirelessly.

This presentation will showcase standards-based multimodal applications that illustrate the principles of the W3C Multimodal Architecture and how the emerging multimodal standards being developed by the group can be applied to integrate speech and GUI technologies and simplify the creation of interoperable multimodal applications.

Openstream has built a suite of healthcare applications incorporating its multimodal SmartCare technology for the use of care-givers, physicians, and patients who are self-monitoring. This healthcare application combines speech and visual images allowing users to type, speak, and scribble on their mobile devices to record vital readings, highlight regions of interest on images and submit them wirelessly. The system stores the images, overlay –gestures, and spoken comments and plays them upon retrieval allowing other experts to review/add their comments and recommendations in a similar way.

While session C303 discusses architectures and languages for developing multimodal applications, this session discusses how to build and test multimodal applications. The first presentation will discuss the pros and cons for four strategies and how to determine the best strategy for your application. The second presentation discusses how to conduct usability studies, a critical step in improving an application’s ease of use.

There are several different strategies for designing multimodal dialogues for use on the handset. Some appear to be less multimodal and more multimedia while others achieve true multimodal interaction. This talk will present different strategies along with the pros and cons for each and how to determine which strategy might be the right one for your application.

As multimodal applications become more widely available, planning for and conducting usability studies is a critical step for improving the ease –of use of these applications. This presentation will discuss strategies for testing the usability of multimodal applications, including how to set up the test environment and some of the special usability concerns a multimodal application presents. Lessons-learned from an on-site multimodal usability study will be included as well.

As more businesses and consumers depend on speech technology, it’s becoming increasingly important to protect your investments. With more than fifteen years of experience in exploring computer security, Kevin Mitnick is a largely self-taught expert in exposing the vulnerabilities of complex operating systems and telecommunications devices. As the world’s most famous (former) hacker, Kevin has been the subject of countless news and magazine articles published throughout the world. Kevin will share how hackers compromise your security and what you can do about it.

Financial services and insurance companies are turning to speech technology for help in meeting stringent security, privacy, and compliance demands. The first case study will show how a financial services company uses a real-time speech solution that recognizes callers who typically call for specific reasons and proactively engages in data-driven dialogues to offer information they usually request. The second case study will highlight how an insurance company provides members with prompt access to information using speech technology, which helped to cut costs, reduce call volumes, and maintain a high level of customer satisfaction.

Wachovia uses a real-time, rules-based, customer-segment-driven decisioning solution that recognizes callers who typically call for specific reasons and proactively engages in data-driven dialogues to offer information they usually request. This approach gives the customer information faster and bolsters IVR containment while simultaneously improving the customer experience. This session will recap the approach, solution, and current results, allowing attendees to learn more about this data-driven decisioning technology and how the IVR can come full circle in providing personal treatment.

Health Net realized cost savings between $2.3 million and $4.4 million annually using speech technology to automate processes, reduce call volumes, and maintain a high level of customer satisfaction. This session will discuss how speech technology can help healthcare providers balance cost effectiveness and service goals, while providing members with prompt access to information and how to enhance security, privacy, and HIPAA compliance using voiceprint identification.

This session addresses two important testing problems. The first is capacity testing. Most businesses have peak periods of high activity throughout the year. Learn how to define and implement testing that will help mitigate the risk of implementing a system that is not ready for peak-period deployment. The second is using open source software for testing. This session will review an integrated, open standards application tuning-based approach based on Eclipse. Application reporting, performance application logs, tuning components, as well as how the latest approaches to integrated tuning tools complement the application tuning process will also be
covered.

Most businesses have peak periods of high activity throughout the year. Making sure that your contact center solution infrastructure and related technologies will work as required under these peak periods is a challenge. In this session the attendees will learn how to define and implement testing that will help mitigate the risk of implementing a system that is not ready for peak period deployment.

Tuning is an absolutely critical step in the success of any speech implementation. Today speech application developers use various application development packages but often rely on separate and costly commercial tools or tool scripts for application tuning. This session will review an integrated open standards based approach based on Eclipse, which simplifies direct correlation between the design of the application and its real-world performance. Application reporting, performance application logs, tuning components, as well as how the latest approaches to integrated tuning tools complement the application tuning process will be covered.

Speech applications are a driving force in defining the customer care experience. Designing and deploying automated speech solutions from the customer’s point of view is a critical factor for competitive advantage. Learn the value from using speech to optimize your customer interactions to drive higher business performance, increase automation rates and improve your customer satisfaction. This session will show the value of understanding your speech applications and delivering a rewarding customer experience.

OrderCatcher is a cost-effective self-service speech application for Chinese and Pizza take-out. Learn whycustomers choose speech self-service, the business problem they wanted speech self-service to solve, how they developed their requirements for the speech self-service application and what advice they would give to a company considering speech self-service to take their fast food orders over the phone.

Financial services firms use speech analytics to identify customer behavior and experience to improve customer treatment and ultimately increase financial benefit. As a result, it can answer these questions: What is actually driving calls into the contact center? Which call types could be handled by self-service? How prevalent are certain agent behaviors? Which agents need additional training? What is driving repeat calls and inappropriate transfers? How can I improve first-call resolution and increase the quality of the customer experience? This insight enables financial services firms to set improvement priorities on facts and to maximize user adoption and customer retention. Allen will discuss how with speech analytics, Capital One was able to achieve quantifiable benefits in the following areas, resulting in multi-million dollar savings: customer experience, sales effectiveness, process improvement, and first call resolution.

Tuning a speech recognition application results in a more accurate solution that can improve the caller experience. The process involves using prompt, grammar, call flow, and caller data to help developers make the needed changes. Two experts will present their favorite tips and techniques for effectively tuning speech applications.

With the help of numerous real-world examples, this session will present a systematic approach to the task of tuning speech applications using field-collected utterances. Areas covered by the tutorial include effective grammar development and maintenance, the process of tuning grammars using field utterances, key metrics to measure grammar performance, techniques to efficiently manage out-of-grammar utterances, and tuning confidence scores and setting confidence thresholds.

While nearly 40 percent to 50 percent of total development and deployment time is often spent on application tuning, there are techniques to create a repeatable process that reduces time and gives the desired results. This presentation will address how to effectively tune speech applications by presenting practical tips and proven techniques.