VUX World

We speak to the creator of the world’s best chatbot about how to design Loabner prize-winning conversational experiences.Steve Worswick is the creator of Mitsuku, the general conversation chatbot that has won the Loabner prizefor the last two year’s straight.13 years in the making, Mitsuku passed the Turing testand convinced a panel of judges that it’s human over the course of a 20 minute conversation, two years in a row, to be crowned the world’s best chatbot and conversational agent.It's featured in the Wall Street Journal, BBC, The Guardianand Wired. And, unlike most chatbots that focus on serving a specific set of use cases, Mitsuku is a general conversational agent. That means you can speak to it about anything.This week's Flash Briefing question is from Brielle Nickoloff of Witlingo: What would an open source voice assistant look like? Send us your thoughtsand you could feature on the VUX World Flash Briefingthis week!What about voice?Although Mitsuku is a text-based chatbot, this episode looks at how to take Steve’s 13 years of experience in creating conversational experiences and apply that to the voice first space.In this episodeThis episode is all about how to design and create a world-leading general conversational experience.We get into detail about how Mitsuku is built (hint: it doesn’t use natural language processing or machine learning like most other conversational AI) and how Natural Language Processing-based conversational agents don’t quite hit the mark.Steve tells us about Mitsuku’s rule-based supervised learning and how that’s leading to better experiences.Despite Mitsuku passing the Turing test, Steve tells us why the Turing test is redundant.We discuss user behaviour and how people treat a general conversational agent, from counselling to romance, bullying to marriage and money worries, and how to be sensitive on those topics.We hear how varied responses can increase engagement. So much so that one person has spent 9 hours talking to Mitsuku!We find out how to deal with pronoun resolution and how to refer back to what was said earlier in the conversation.We uncover how brands are using Mitsuku as part of their conversational experiences, handing off to her when a user strays away from the use cases that their bot can handle.We chat about how Alexa fairs against Mitsuku and hear where Siri would have finished if it was entered in to the Loabner prize competition.Perhaps one of the most valuable lessons in this episode is the importance of persisting. Creating a conversational agent, a true conversational experience, takes time. It’s not a quick fix that you cobble together with a quick Alexa Skill. It takes years of development, iteration and constant improvement. But, if you stick with it, you might end up with the next best conversational agent.Our guestSteve Worswick started out in IT support and built Mitsuku as a passion project on the side. 13 years of hard work and 3 Loabner prizes later, he’s now working at the world’s largest chatbot agency and provider, PandoraBots.Where to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play VUX World on AlexaLinksContact PandorabotsCheck out Mitsuku on PandorabotsTalk to MitsukuCheck out Steve's talk at the Chatbots and Voice Assistants Londonevent Hosted on Acast. See acast.com/privacy for more information.

This week, we're finding out how brands can get started and enter the voice first world of smart speakers and digital assistants.Me and Dustin Coates are joined by one of the top US voice first agencies, Witlingo. We speak with two Lead VUX designers, Luciana Morias and Brielle Nickoloff, about how your brand can bridge the gap over to voice.In this episodeBrielle and Luciana share how they guide brands through the process of discovering their voice and establishing a voice first presence.We discuss the new challenge of working out what your brand sounds like and how to determine whether to focus on voice first content or voice as a service.They discuss how brands should be playing the long game and the challenge of convincing clients to start small and adopt a continuous improvement culture to grow their voice first capability.We chat about figuring out whether your should repurpose existing content or create new and discuss some of the great guides to voice design that Witlingo produce, including the guide to making your Facebook content voice friendly.Our guestsLuciana Morais has a background in UX research and analysis and has a wealth of design experience. Now working at Witlingo as UX Lead and VUI Designer.Brielle Nickoloff has a background in linguistics and has published a study on The use of profane threats and insults in the Anthropomorphization of digital voice assistants. Brielle is also Lead Voice User Experience Research and Design at Witlingo.Where to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play VUX World on AlexaLinksVisit the Witlingo websiteFollow Witlingo on TwitterRead Witlingo's VUI assessment guidelinesRead Witlingo's Facebook guidelinesFollow Brielle on TwitterFollow Luciana on TwitterCheck out the Ubiquitous Voice SocietyRead Brielle's paper: The use of profane threats and insults in the Anthropomorphization of digital voice assistantsIt's about the interface stupid Hosted on Acast. See acast.com/privacy for more information.

This week, me and Dustin are speaking with the Fresh Prince of AI, Frank Schneider, about how Speakeasy AI aims to deliver the promise of AI in voice (that’s a lot of AI’s).How many people truly understand what their customers are asking for? Whether it’s in your Alexa Skill, your chatbot or in your IVR, you can’t hope to serve the needs of your users or customers if you don’t understand what they’re trying to do or ask.Understanding is the most important first step you can takeOnce you truly understand the current situation, you can realise whether you’re meeting your existing customer needs, and how well you’re doing that.Through gathering understanding, you can also work out where you’re failing and where the opportunities for improvement or expansion are.That then helps you improve and plan for the future.Speakeasy AI is helping businesses understand what their customers are trying to accomplish on a wide variety of conversational platforms by extracting the intent from any conversation.Its patent-pending technology, called Speech-to-Intent, doesn’t use the typical speech-to-text engine that most voice-first platforms use. Instead, it analyses the actual audio in real time through funnelling it through a pipeline of different ‘top secret’ micro services.This means that low audio quality and accents have no effect on its ability to understand customer intent. Plus, it also allows for further understanding of context.In this episodeDustin Coates and I hear from Frank Schneider, CEO, Speakeasy AI, about the current state of play in the AI field and touch on the amount of bullshit that exists right now.We discuss how conversational understanding works and why speech-to-text might not be the most optimum way to capture intent.We delve into the ins and outs of Speakeasy AI and get the low-down on its patent-pending Speech-to-Intent technology and hear how it could be a better way of understanding customer intents, regardless of audio quality or accents.Frank tells us all about how Speakeasy AI can help businesses improve any conversational platform. He shares the opportunities that exist in the IVR space and how much untapped potential there is for businesses who’re willing to listen.We've discussed VUI design for IVR with Simonie Wilsonrecently, and it would seem that you could use Speakeasy AI as part of a discovery piece of work to figure out where to start, then use Simonie's techniques to begin making improvements.We also chat about the challenges of the AI industry and how working together could bring progress.Our guestFrank was born and raised in Philly and, after spending 9 years in education, including teaching at a school for high school kids who committed felonies, he transitioned into technology sales and marketing, where he’s spent the last 13 years.He’s consulted and led teams providing solutions in various SaaS and AI solutions for contact centers and B2B. He was the first sales executive at Creative Virtual USA and helped grow the team from 12 to 40 employees. After a successful exit, his former CEO is now funding his new venture, Speakeasy AI.Where to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play VUX World on AlexaLinksVisit the Speakeasy AI websiteFollow Speakeasy AI on Twitter Hosted on Acast. See acast.com/privacy for more information.

This week, we’re finding out how content creators can have their podcasts and YouTube content indexed and searchable on voice, with Bryan Colligan of Alpha Voice.With the podcast industry thriving and more people listening to podcasts than ever, more brands are starting to launch their own podcasts. Podcasts are a perfect fit for devices like the Echo and Google Home because they provide ambient entertainment, similar to the widely popular relaxation sounds skills.Two problems face podcast and content creators: how do you make your podcast discoverable in the first place and how do you allow people to search through your backlog of episodes in order to find something that interests them?Podcast discoverability is almost as much of a problem as Alexa Skill discoverability. Although Google is beginning to do its bitto help podcasts be discovered online, what about on voice?This is the problem Alpha Voice aims to solve.Help others get their skill passed first time by sharing your skill certification stories: Send us your tipsand you could feature on the VUX World Flash Briefingthis week!What is Alpha Voice?Alpha Voiceindexes your podcast or YouTube content and makes it all searchable on Alexa via your own Alexa Skill.And it’s not just the podcast titles and guests you can search for. You can search for anything at all that interests you and the platform will search within your content to find your search term, then recommend that episode for you to listen to.In this episodeWe’re talking to Alpha Voice co-founder, Bryan Colligan, about how the platform works, how he and his co-founder built it and what value it gives content creators.We also get into detail about how the VUX of search works on voice: processing and serving potentially hundreds of search results. How do you determine which ones to display to the user?We also discuss:The 5 ways to monetise contentSkill certification inconsistencies, including censorship and 'unwritten rules’How you can get up and running with Alpha VoiceWe wrap up by telling you all about the VUX World Alexa Skill, built using Alpha Voice! (U.S. only right now but will be available in EU soon.)Our guestBryan Colligan is an entrepreneur and the co-founder of Alpha Voice. Bryan is based in Silicon Valley, has founded a series of startups and has been helping startups create mobile apps and improve their SEO for the last 10 years.After reading the Mary Meeker internet trends report and learning that Google can understand 96% of what humans say, Bryan has turned his attention to the voice-first world.After a number of failed experiments, he stumbled across the idea for Alpha Voice and is now helping content creators have their content found on Alexa.Where to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play VUX World on AlexaLinksAlpha Voice YouTubeVisit the Alpha Voice websiteFollow Alpha Voice on TwitterFollow Bryan on TwitterConnect with Bryan on LinkedIn Hosted on Acast. See acast.com/privacy for more information.

This week, we’re getting deep into voice analytics and will help you learn more about how you can understand the performance of your voice first experience.One of the biggest benefits that technology has given us is the ability to understand. To understand whether our latest PPC campaign had an impact on sales. To understand whether our new website increased our leads. To understand whether our pricing tweak made a difference on click through rates. To understand whether our foray into Facebook is sending more traffic. To understand whether our customers are satisfied.Tools such as Google Analytics have been providing this kind of value to website owners for years. Tracking where your users come from (Google, Facebook etc), what they do when they arrive and whether they convert are the cornerstones of understanding website performance.What about voice analytics?With the introduction of new mediums such as conversational chatbots and voice first applications on platforms such as Alexa and GoogleAssistant, how do you understand the performance of these things?How do you know if your Alexa Skill or Google Action is successful? Send us your answersand you could feature on the VUX World Flash Briefingthis week!Can you apply the same rules as the web? Can you even access the same data? Is there some new metrics that matter more? And how can you use all of this to understand and improve the performance and use of your product?Well, that’s what you’re about to find out.In this episodeWe’re speaking to Dashbot.ioCEO Arte Merritt all about the conversational analytics platform and how you can understand whether your conversational experience is working for your users.We discuss the kind of metrics Dashbot provide including:No. usersRepeat usersTime per sessionRetentionSentiment analysisMessage funnelsIntent funnelsTop exit messagesAI performanceGoalsBehaviour flowConversation flowArte tells us some case studies of how the tool has been used to understand and then improve conversational experiences.We discuss some of the challenges with conversational analytics and how they relate to the voice first space and we hear about where voice analytics are heading in the future.Our guestArte Merritt has worked in mobile and analytics for 20 years. He built an analytics platform which he sold it to Nokia before turning his attention to fill a gap in the market when he realised that Slack didn’t have any analytics. Dashbot was born and its been serving conversational designers ever since, helping them understand and improve their chatbots and voice applications. Since its creation, Dashbot has analysed 32 billion messages and counting!Where to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play V.U.X. World on AlexaLinksCheck out Dashbot.ioFollow Dashbot on TwitterFollow Arte on TwitterAlexa Skills: competitive intelligence Check out the Dashbot BlogFind out more about the Smart Voice Summit London Hosted on Acast. See acast.com/privacy for more information.

This week, we take a look at the similarities between VUI design for IVR and VUI design for voice assistants. We also explain what VUI tuning is and why it’s important, whilst giving you some tips on how you can tune your voice user interface. We also discuss PinDrop and voice first security.In this episodeWe speak to one of the world’s expert VUI practitioners, Simonie Wilson, to get under the hood of Passport and figure out what it is, how it works, why it’s needed and how you can use it to authenticate users with confidence whilst preventing fraud.We also tap into Simonie’s vast VUI design experience and discuss how she goes about designing VUIs that delight rather than smite customers. We get into detail about the benefits of VUI tuning and Simonie shares her advice on how you can continuously improve a VUI experience.Are brands failing on Amazon Alexa and Google Assistant? Send us your answers and you could feature on the VUX World Flash Briefing this week!Privacy and securityPrivacy is often cited as a barrier and a challenge in the voice first space. How do you authenticate a user, build trust and enable people to transact in a frictionless way, all without a long, drawn out, failure stricken on-boarding process?PinDrop is changing that with it’s product, Passport: a fool-proof way to recognise whether someone is who they say they are simply by the sound of their voice. It works in all voice first areas and can even tell whether the voice is synthetic.Here's an example of it working with Alexa:There is so much potential in the voice first space from a vcommerce, health and financial management perspective that technology such as this could smooth over the cracks in the verification process and enable people to transact more seamlessly in a voice first world.Our GuestSimonie Wilson is the queen of VUI design. With over 20 years experience working in the speech and VUI design space, Simonie's career has included working with large companies such as Microsoft and GM, small companies such as startups and contracting too. Simonie has knowledge and experience in the VUI design space that few others do and is one of the few people to have extensive experience with VUI tuning.Simonie is madly passionate about VUI design and, in this episode, shares all of that passion and some real lessons and insights from her experience that’ll help all VUI designers improve what they do.Where to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play V.U.X. World on AlexaLinksVisit the PinDrop websiteCheck out PinDrop on YouTubePinDrop on FacebookPinDrop on TwitterConnect with Simonie on LinkedInEmail SimonieRead about PinDrop Passport in Forbes Hosted on Acast. See acast.com/privacy for more information.

This week, we’re digging into how you can create an Alexa Skill using BotTalk and we give you a template for running a voice first discovery workshop, with SmartHaus Technologies CEO and BotTalk co-founder, Andrey Esaulov.We discuss the importance of starting with a solid use case and how imperative it is to base your voice app on a real-world scenario that’ll add value to your users.What turns an average voice experience into an EPIC voice experience? Send us your answers and you could feature on the VUX World Flash Briefing this week!We then dive deep into the practical detail of how to approach designing a voice first user experience with BotTalk and find out more about the language it’s built in: YAML. We discuss what BotTalk is, how it’s different from some of the other tools on the market, how it works, it’s features and how you can get up and running.Finally, Andrey takes us through a voice first discovery workshop template that he uses with clients in order to take a brand from zero to hero: from ideation to prototype, and how you can do the same too.We also traverse some other interesting conversational landscapes such as the concept of skill-first companies: brands that pop up as skills which are the core of the business, like an app is for Instagram. We chat about Artificial Intelligence and how intelligent it actually is in the voice first space. We touch on managing client expectations, monetisation and how voice is making waves in Germany.About BotTalkThe current selection skill building tools on the market are at opposite ends of the technical spectrum. Some tools require you to know how to code from the ground-up, like Jovo and be a skilled back-end developer. Others have a drag and drop interface and don’t require any coding at all, like Storyline.BotTalk bridges the gap between those two worlds with a tool that’s aimed at UX designers who have some basic coding knowledge, like HTML and CSS. It provides some of the technical capability you’d expect if you built something from scratch, whilst providing a more simple coding language: YAML. Think of it as HTML for voice.Our GuestAndrey Esaulov is the CEO of SmartHaus Technologies, which specialise in growth hacking in the mobile space, and the co-founder of BotTalk, a voice first and bot application building platform.Andrey has a computer science background, with expensive experience in the start up world and mobile growth space, as well as a PhD in Linguistics and Literacy.Andrey’s skillset is a perfect match for this industry and his knowledge in this area is vast. Couple his computer science and linguistics knowledge with his skills in working with clients and delivering growth and you’ve got a perfect recipe for success.LinksCheck out BotTalkFollow Andrey on TwitterJoin the BotTalk Facebook communityFollow BotTalk on InstaWatch the BotTalk tutorials on YouTubeVisit the Smarthaus Technologies websiteJoin the Alexa Slack channelEnable the VUX World Flash BriefingFeature on this week's Flash BriefingWhere to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play V.U.X. World on Alexa Hosted on Acast. See acast.com/privacy for more information.

Dr. Pete, Marketing Scientist at Moz, and world-leading SEO oracle, tells all about the voice search landscape, and how you can rank for searches on digital assistants like Google Assistant and Amazon Alexa.This is a jam-packed episode with deep, deep insights, advice and guidance on all things voice search related. We'll give you practical ways to compete to be the answer that’s read out in voice first searches, as well as some notions on the current and potential future benefit that could bring.Voice searchThere are all kinds of stats around voice search, which we’ve touched upon before.Gartner predicts that 50% searches will be voice based by 2020There are already over 1bn voice searches performed per monthWith more people using their voice to search, how will that affect search marketers, content creators and brands?What’s the difference between a voice search and a typed search?Is there anything you can do to appear in voice search results?We speak to one of the search industry's top sources of SEO knowledge, Dr. Pete, to find out.Getting deep into voice searchIn this episode, we’re discussing the differences between voice search on mobile, voice first search on smart speakers and typed search.We discuss the absence of search engine results pages (SERPs) in a voice first environment and increased competition for the singularity: the top spot in voice search.We chat about the search landscape, the effect voice is having on search, changing user behaviour and expectations, new search use cases and multi modal implications, challenges and opportunities.We get into detail about how voice search works on devices such as Google Assistant and Google Home. This includes debating Google’s knowledge graph and it’s advantages and disadvantages in a voice first context.We look at the practicalities of serving search results via voice. This touches on the different types of search results, such as featured snippets, and how voice handles different data formats such as tables. We get into detail about the different types of featured snippets available and how each translate to work (or not work) on voice.We discuss Dr. Pete’s work and studies in the voice first space including his piece ‘What I learned from 1,000 voice searches' and what he found.We wrap up with some practical tips that you can use right now to start preparing for the influx of voice searches that’ll be hitting the air waves soon and help you start to rank in a voice first environment.Our GuestDr. Pete Myers (a.k.a Dr. Pete a.k.a. the Oracle) is the Marketing Scientist at Moz, the SEO giant and search industry leader.Dr. Pete has been an influential search marketer since 2012 and has spent years studying Google’s search algorithm, advising clients and the SEO industry on best practice and guiding the industry into the future.His research and writing on the topic has been helping brands keep on top of the search space, improve their rankings and business performance and has helped keep Moz at the top of the industry.Mozhas been at the top of the SEO chain since 2004 and is trusted by the whole SEO industry as the place to go for SEO tooling, insights and practical guidance.LinksFollow Dr. Pete on TwitterFollow Moz on TwitterRead Dr. Pete's 'What I learned from 1,000 voice searches on Google Home'Read Dr. Pete's work at MozWhere to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play V.U.X. World on Alexa Hosted on Acast. See acast.com/privacy for more information.

We’ve covered plenty of voice first designand developmenton this podcast. Well, that’s what the podcast is, so we’re bound to! Most of what we’ve discussed has largely been voice assistant or smart speaker-focused. We haven’t covered a huge amount of voice first application in the browser and on mobile, until now.Mic checkYou’ll have noticed the little mic symbol popping up on a number of websites lately. It’s in the Google search bar, it’s on websites such as EchoSim and Spotify are trialing it too. When you press that mic symbol, it enables your mic on whatever device you’re using and lets you speak your search term.Next time you see that mic, you could be looking at the entry point to Voysis.On a lot of websites, that search may well just use the website’s standard search tool to perform the search. With Voysis, its engine will perform the search for you using its voice tech stack.That means that you can perform more elaborate searches that most search engines would struggle with. For example:“Show me Nike Air Max trainers, size 8, in black, under $150”Most search engines would freak out at this, but not Voysis. That’s what it does.Of course, it’s more than an ecommerce search tool, as we’ll find out during this episode.In this episodeWe discuss how approaches to new technology seem to wrongly follow a reincarnation route. Turning print into web by using the same principles that govern print. Turning online into mobile by using the same principles that govern the web. Then taking the practices and principles of GUI and transferring that to VUI. We touch on why moving you app to voice is the wrong approach.We also discuss:Voysis - what it is and what it doesGetting sophisticated with searchesDesigning purely for voice vs multi modalThe challenge of ecommerce with a zero UIThe nuance between the GUI assistant and voice only assistantsHow multi modal voice experiences can help the shopping experienceMaking the transition from GUI to VUIThe similarities between moving from web to mobile and from mobile to voice - (when moving to mobile, you had to think about gestures and smaller screens)Error states and points of delightThe difference between designing for voice and designing for a screenTesting for voiceUnderstand voice first ergonomicsOur GuestBrian Colcord, VP of Design at Voysis, is a world-leading designer, cool, calm and collected speaker and passionate sneaker head.After designing the early versions of the JoinMe brand markings and UI, he was recruited by LogMeIn and went on to be one of the first designers to work on the Apple Watch prior to its release.Brian has made the transition from GUI to VUI design and shares with us his passion for voice, how he made the transition, what he learned and how you can do it too.About VoysisVoysis is a Dublin-based voice technology company that believes voice interactions can be as natural as human ones and are working intently to give brands the capability to have natural language interactions with customers.LinksCheck out the Voysis websiteFollow Voysis on TwitterRead the Voysis blogJoin Brian on LinkedInFollow Brian on TwitterListen to the AI in industry podcast with Voysis CEO, Peter CahillRead Brian's post, You're already a voice designer, you just don't know it yetWhere to listeniTunes/Apple podcastsSpotifyStitcherTuneIniHeartRadioPippaAny other podcast player you use or ask Any Pod to play V.U.X. World on Alexa Hosted on Acast. See acast.com/privacy for more information.

Voice first games are one of the most popular Amazon Alexa skill categories. So what type of voice games are available? And how do you create them? We speak to game developer and reviewer, Florian Hollandt, to find out.Games are helping Alexa take off. According to Voicebot.ai, Alexa Skill games are the second most popular skill categorybehind smart home skills. Amazon has been encouraging the development of games, too. We've seen the Alexa Skills Challenge: Kids recently and I'd say it’s more than likely that most of the developer rewards will have gone to game developers, given the engaging nature of games.We’ve touched upon voice first games on the podcast previously, such as our chat with Jo Jaquinta of Tsa Tsa Tzu, but we haven’t yet covered audio game development in detail, which is what we’ll do today.Creating voice first gamesIn this episode, we’ll be getting into detail about the different kids of voice first games that are out there, as well as looking at some of the techniques you can use to create engaging games such as interactive stories.We’ll cover things like:Naming a game and how a name can reduce discoverabilityThe challenge of providing contentThe one game per month challengeThe types of games that are available on Amazon AlexaGame design techniquesInteractive story game development techniquesFake decisions - what are they and how can you use them to enhance engagementOur GuestFlorian Hollandt is the Product Manager at Jovo, the cross platform voice app platform, and is also an Alexa game developer and reviewer. He’s created some popular games on Alexa, such as the German card game, Mau Mau, and has written a ton of voice first game reviews on Medium.Florian is madly passionate about voice first games and his knowledge on the subject is impressive. He guides us through his experience and shares some delightful tips on how you can start creating voice first games yourself.LinksFollow Florian on TwitterCheck out Florian's voice first game reviews on MediumSome of the things Florian spoke about:Voice first game design tutorial from AmazonThe Mitsuku chat botWoebotOne game per month websiteSelect a storyMagic doorMr Robot Hosted on Acast. See acast.com/privacy for more information.