Speaker

Dan Fitzpatrick

Dan Fitzpatrick

General Manager at Frank REPLY, Practice Lead for Voice Machine Interfaces, REPLY AG

Frankfurt am Main, Germany

Although my academic background was in language studies, I’ve been working with internet technologies for over twenty years, mostly doing classic web programming, followed by leading teams of developers. Technically, I have always been passionate about interface technologies. Now, I can combine my deep interest in linguistics and voice interfaces to contribute to the voice revolution.

I firmly believe that we, as consumers, are part of the “Voice Revolution”. Use cases, market potential and technological breakthroughs are growing exponentially. Because we have all the necessary building blocks at Reply, we need to document, bundle and channel our efforts to provide excellent voice solutions to Reply customers – and indeed – to ourselves, the IT-consumers of the 21st century!

Area of Expertise

  • Information & Communications Technology
  • Finance & Banking
  • Media & Information
  • Travel & Tourism

Recipient Design: the Science of Conversation

The synthesis of conversational analysis and NLP / NLU applications (combined with the power of large language models) has immense potential. Foreseeably, there are several areas of activity and expertise necessary for the implementation of both visual and conversational interfaces which will face radical disruption if not complete obsolescence. These are: the inclusion of emotion AI and sentiment analysis as building blocks of service-driven architectures, the strategy of audience design and, lastly, the documentational development of persona and tonality as static UX “guardrails” and, of course traditional, flow-driven, conversational design.

Even though the science and field of studies for recipient design are not new, they has yet to be combined with conversational design using modern generative AI technologies. This session will lay the groundwork for a new interpretation of the practice of hyperpersonalisation and an entirely new aspect of conversational architecture: prompt compiling – based on the principles of conversational analysis and recipient design.

Chinese Rooms and the Wizard of Oz

If you’re curious as to how the psychological and philosophical paradigms postulated by thinkers in the 1980s have greatly contributed to modern, systematic interface development, this is the webinar for you.

Find out how to improve iterative, automated testing routines with human intervention to increase the speed of development. Testers are led to believe that they are interacting with software functions, such as language processing. In reality, the functional features are being simulated by humans to rapidly fine-tune the interface.

This testing method is particularly effective for voice interfaces, because test supervisors can easily simulate and adapt machine responses and intuitively interpret test candidate inputs which might otherwise be misinterpreted by the interface itself, resulting in system failure.

The moderators start with a humorous, anecdotal historic background of the two theories revolving around Chinese Rooms and the Wizard of OZ.

Subsequently, they build the bridge to modern interface development.

Participants can learn first-hand and interactively about the challenges and solutions to successful interface design and testing – based on principles over forty years old.

Getting Intimate with Alexa – Helping Nerds Get Dates!

Are you a nerd? Then this workshop might be for you! Anyone who’s ever watched The Big Bang Theory knows that nerds often have problems getting dates. The solution: use technology!

What I’ll talk about:
After starting with a quick nerd test, we’ll deep-dive into speech processing technologies such as NLU, NLP and ASR to get a quick overview of the terminology, examining the challenges along the way.
Then we’ll go into the four best practices for developing voice interfaces, giving all the nerds or romantically challenged members of the audience the tools they need to successfully ask that special someone ❤ out on a date.

Find out how (or how not) to:

* set up the time and place
* order dinner for your date
* keep the conversation moving forward
* not to sound like a wooden broomhandle!

Where my passion comes from
After my studies in foreign languages and language science, I’ve maintained an affinity for linguistics all my life. Working as a programmer, it was natural to develop an interest in the technologies which enable humans and computers to communicate using voice interfaces.

Why you should attend
The proliferation of assistants in the past few years has promoted rapid development of language processing technologies, as well as highlighting the challenges in properly programming the interfaces in the applications. Come find out what to do (and what to avoid) when programming for speech… …or trying to land a date!

Discount Psychology with Alexa – Playing with Sentiment Analysis

In this entertaining and interactive session, we’ll have fun with one particular AI technology – sentiment analysis – and see how it can be used to process human statements, such as movie, product or book reviews, or categorize the tone of customer-service calls. You’ll not only get an interactive, hands-on demonstration of voice interfaces and sentiment analysis, but you can also have your joys and deepest fears analyzed for free by a panel of famous psychologists.
Tech-Tags: #alexa #lambda #nodejs #amazon+comprehend

Who am I? A Vision for Voice Authentication

This entertaining session will introduce participants to the world of voice authentication. With the rapid rise in acceptance of voice interfaces, it's inevitable that the underlying applications will have functions requiring secure communications. In addition to the usual challenges posed to system architects when designing a secure environment, voice interfaces offer new hurdles. Entering a PIN number or typing in a password are events which can be executed with a level of relative secrecy and are unequivocal – being entered either correctly or incorrectly.

What would you do on the phone, in a crowded room or at the supermarket if you were forced to shout out your password as a form of authentication? Nothing at all, in all probability! There is however increasing demand for secure voice authentication – and the methods and technologies involved are becoming more reliable. Hands-free operations are the way of the future, let’s find out how to make them secure.

The sessions includes a Demo with Amazon Alexa

Dan Fitzpatrick

General Manager at Frank REPLY, Practice Lead for Voice Machine Interfaces, REPLY AG

Frankfurt am Main, Germany