National Analytics Conference Ireland, and AWS User group.

National Analytics Conference Ireland, and AWS User group.

Talking about the search for a new data person, story telling and Terraform.

This episode I talk about attending the National Analytics Conference Ireland where I'm looking for a new data science/ML/AI type, and an AWS user group meetup where I fell in love with Terraform.

Tue, 21 Nov 2017 09:10:00 GMT
duration: 34:07
Spotify | iTunes | Sticher | Google Play | Player.fm | MyTuner Radio  

Transcript:

Hello and welcome to the podcast. I’m your host, Dave Albert. In this show, I talk about technology, building a company as a CTO and co-founder, and have guests to discuss their roles in technology and entrepreneurship.

Hello, folks. Welcome to another episode. I remembered the word this time of the podcast. This episode, how about - alright, I’m going to leave that in there. I said I was going to go raw, so I’m really going to go raw. If I make mistakes, that’s what happens which is the case in real life.

So today - today - this episode, I say, wow, I’ve got to be better at this, don’t I? Please don’t run away yet. I’m going to tell you about yesterday. So yesterday, I went to an interesting conference and an interesting meet up. The conference was the National Analytics Conference. So in the last episode, I mentioned that we are losing our current data person, so I went to this analytics and data conference hoping to find some connections or possibly somebody who could be an employee, or maybe a service that could help us out while we are in the process of trying to find a replacement. We’ve got someone working with us a little bit as an intern but hopefully, that turns into a real thing, a nice full-time team member. We’ll see. The conference was pretty interesting. It was a good mix of some - I wouldn’t say startups, so they mentioned that startups but some enterprise, some government agencies, and some consultant types who were used to working with pretty large enterprises.

The first speaker was Chris Williams, the Chief Architect of Watson at IBM. I believe that’s in Europe, probably not worldwide but I could be wrong. I can’t remember. I remember at least one of the speakers at one of the events I went to yesterday that had chief-something and it was for Europe, but it was not mentioned that it was Europe. He was talking about obviously, Watson and said that AI is really Augmented Intelligence instead of Artificial Intelligence. I thought that was pretty interesting. I think we all kind of know that intuitively but the artificial intelligence seems like it’s a little more scary, the machines taking over whereas the augmented intelligence is what it really is which is the same as technology. Technology just augments our life; it doesn’t replace what we do. Well, I guess technically, it does replace what some people do, but it hasn’t replaced people. It may have replaced some of the tasks that people do. So computers are never going to think for us. They’re going to assist us in thinking. So I found that pretty interesting, showing how the AI and machine learning can be used to improve people’s lives. Our product Medit that we’re in the process of building, we’re using machine learning and AI to help sift through the massive amounts of information for the healthcare providers, so we want to find the most relevant information, and computers are pretty good at identifying patterns, and so if we can identify what levers to look at - see, I am not the data person; I’m the tech guy. I can deploy things, I can write code, but using the right word sometimes escape me. So the right elements - gosh, I can’t think of what that’s called - anyway, if you look at the right vectors - yes, I believe that’s the word I’m looking for - if you look at the right vectors, a computer can categorize and identify matching elements and deliver that to the user so that they are not sifting through piles of junk or things that they’re not interested in which it would be great if I could get some AI to scan the through social media and remove all the politics. I don’t have that person’s name. I’m actually going to go and grab the paper so that I can say who this is. Hang on. And I’m back. Sorry about that. So Rob Krizman, the Data and Analytics Advisory Leader for EY in the UK and Ireland. He had a pretty interesting slide that a common question that he gets is, “Why can’t I see it?” which is extremely relevant to what we are doing at the moment. So even if you track everything, if you can’t find it or see it, what’s the value? Which that is so relevant to us because we’re not 100% sure of all the elements that we’ll want to track and analyze, so we have to track everything, but if everything gets tracked and we don’t have a good way to figure out what we need to care about, then what’s the point? So we need to move from insights to action, and then to value, so you can’t forget about the value. It’s just like I like to remember about agile, that it’s not about doing work faster, it’s about delivering value to your users and customers more quickly.

The next point was from Yvonne Holmes, the head of Business Performance and Analytics at AIB which is a pretty large bank in Ireland, and she said that people don’t remember facts; they remember a story. So she showed us her own story moving towards kind of a transformation within their large organization to use analytics to deliver more value for the business. So as an example, before telling her story - well, her team’s story - she showed some slides about the storytelling from Barack Obama where it was about his Hope and Change speech. It begins with evoking an emotion such as pride, and then you revoke that emotion, so you show how it’s not as good as it could be, and then show how you’re going to make it better, so refrain. So he evoked the pride, where he came from, and then revoked it by explaining how poor the education system is, and then reframed it with hope plus action, so I don’t know if anyone has seen the [9:34] describing the different types of stories and that step one is it has to go up and then it has to go down, and then it comes back up, and so I guess that’s the typical redemption story. So it’s all about the emotion that people find, so you need to use that in telling your own story. So whether that’s telling it to your customers and your users or telling it to the people who work for or with you, that is how you get people to buy into what it is you’re trying to sell. I mean, and not just sell as in trade for money but to sell the idea or to get buy-in from your team. I think we all have heard at one point or another that it’s about the story, but I thought that the way she described it was a pretty good reminder of how you need to structure your story to make sure that it has those elements of ups and downs.

Then, during a panel discussion, there was a really interesting quote that a gentleman gave, it was there is no such things as products or service; it’s all stories, which I think that’s really neat. It’s about the story your customers, clients buy into about you or about the future you’re offering or bringing them. It’s not about features. The same way that you often hear, it’s about the problems you’ve solved for them, not the features that you bring them. It’s about who they are when they use your products or service, so it’s the story that gets told, whether from you to them or that they tell themselves. We’ve all seen that in our own lives. How many times have you wanted some toy - a toy, I mean, an electronic tool - a phone or a laptop, tablet, a video game system, motorcycle, whatever it is, and before you get it, you have a story of what it’s going to do for you, and sometimes it’s true, and sometimes it’s just another thing that becomes clutter around you. So it’s about that story. It’s not about the actual physical object that you try to acquire.

So that was the first section of the conference which was turning insights into action, data analytics and organizational strategy. So I found several - several points were pretty thought-provoking and I actually had a few personal things that I wrote down to focus on in my own life, not as related to data as you would think and more about how I just want to live myself, live my own life as opposed to specifically looking for the information on data and finding someone for our team. The second session was the theme of the emerging business and technology landscape. That session, I’m not sure that I did or didn’t care for any of it but I seemed to have not written down anywhere near as many notes, so I don’t know if maybe I was getting tired by the end of the day or - I’d also met a person who does sourcing of data, AI, ML type people and knew that I had a solution for part-time - a partial solution to help us through some of the rougher patches where we don’t have the expertise on the machine learning so that we could get over some humps, so it’s potential that I had started to focus on thinking about how we were going to do those elements of improving our app as opposed to really hearing some of what was spoken about at that second session, but also, they seem to be a little bit more above - or not above - they seemed to be more focused at larger organizations but there was one thing that I wrote down - or actually, posted on Instagram, the - who was this? This was - that was the… Caroll. I don’t seem to have his first name. Ah, James Carroll. He is the group head of Consulting in MasterCard. So here in Ireland, MasterCard seems to have quite an interesting group of developers working on some solutions. So one of his quotes through the day was to anticipate what people will want next. Stop looking at your customers and look at other businesses. So I’m not sure if I agree with that or not. [coughing] Pardon me. So I get an idea of what they’re saying, just the same way that if - what has Andrew Forward said, that if he had asked his customers what they wanted, that they’d have one of the faster horse and the smartphone would never have been invented because they didn’t exist before they existed. Well, technically, I mean, Palm Pilot, I suppose was kind of a smartphone. In fact, at one point, it was a smartphone because there was that one Palm Pilot that had a cellphone built into it, but the idea is not - if you’re going to innovate, you may need to focus on what other verticals are doing and try to adapt some of their thinking to what you’re doing. He gave an example but - so yeah, the example was that larger companies that are more tech-focused, like Amazon, Alibaba, a few other - probably Apple - that if you take what they’re doing, that a slower-moving industry like banking and finance can apply those same techniques and learnings and leapfrog banking businesses as opposed to asking what customers want, which I think can kind of be related to - for a bank, it’s so painful to move from one bank to another that you’ll put up with worse products, and by products, I mean tools, like their websites and their mobile apps, because the hassle of changing all your account information at all the different places that touch your account is so - that threshold of pain is so high that you just can’t imagine having to go to a new bank just for a possibly slightly better online experience. So I can see it, but also, don’t obsess over your competition, obsess over your customers. So I don’t know, it’s an interesting one. I’d love to hear what people think. I mean, anytime you want to reach out to me on Twitter or Facebook, or email, I’d love to hear from you. Questions, comments, if you want me to talk about something else or shut up, or whatever, I just love to hear what you have to say.

So that seems to be all I’ve got on the analytics conference. Then, later in the day, after I actually got some code written and some poor requests taken care of, I made it to an A to B west meet up and that was really, really useful. So the first speaker was - I can’t remember how he pronounced his name - Yevgeniy - he goes by Jim anyway, Brikman. He’s the author of a couple of books on Terraform and startups. He’s a co-founder of a company called gruntwork.io and he spoke about Terraform and using Terraform modules to make your infrastructure components reusable. I can’t wait to dig into Terraform. A lot of our - well, yeah, a lot of our automation is built in Ansible and I really do enjoy Ansible and it does what we need, but Terraform does some really interesting things. You know, I always like HashiCorp products but Terraform just looks so good and it’s really nice that its convention includes testing which I think most devops tools are neglecting. So I want to give it a go, probably over the weekend when I’ve got some extra time. Haha, yeah, when you’re in a startup, there’s no such thing as extra time. So Terraform not only can run the bootstrapping and automations of installing software, but it can also interface with your cloud provider to spin up those hosts. I mean, Kubernetes can do that, I’m sure there’s ways for Ansible to do the same, but Terraform, the testing built-in as a convention, the amount of libraries available already look pretty compelling. So I’m going to take a look at it and see if it’s worth trying to move or possibly in conjunction with Ansible since those playbooks are already written. So I don’t know, I’ll let you know when I’ve had a chance to deal with it and see what’s what, and keep you updated.

The next speaker was Hitor Lisa, he’s a Solutions Architect from AWS talking about serverless, so AWS lambdas. He was directly describing a reason that I haven’t used lambdas as a production solution so far, and that when you move to a lambda or serverless to not basically try to build a monolith. So I wouldn’t say that I was necessarily trying to build a monolith, but I think I was building endpoints that were too large, so instead of potentially one endpoint per lambda, I was basically trying to replicate a large set of rust. So for example, posts. So posts would be the articles that get posted that we do, that we deliver within Medit, and I was thinking to basically build posts, all that gets that puts posts as one lambda and that’s not the right way to do it which would I do is build more a constellation of services, is what he called it, and he used step functions so that each step that needs to be done is its own function, and that makes it seem a little more interesting. I really want to hear what people who’ve used lambdas in production have to say, what was good, what was bad, there seems to be some more testing tools available now than there were when I first look into the lambdas, so that makes me a bit more comfortable, and coming soon are blue-green deployments. So that is really nice, and if you’re not familiar with blue-green deployments, that’s where you have one version, you call it blue, then you’re able to deploy another version, green, and bring down blue, but then if green doesn’t work, you can take it offline and bring back up blue instantly, and that way you don’t have excessive downtime. I do that with our current application servers and using Docker containers. It can actually run side-by-side, so you can do - if you have enough blues or enough greens, you could have a Canary testing, so one out of ten to be the new version and check your analytics to validate that you aren’t getting unexpected results. So lambdas may be the future, but I’m still not 100% convinced, but very possible.

The next speaker was, interesting enough, the product looked nice but it looks more enterprise-y than I was interested in, though he did had one pretty funny line where he said that because he was starting his live demo - he was from Australia - that they called that brown trouser time. That gave me a pretty good laugh.

Who else did we have? We had Ian Massingham, AWS Worldwide Lead and Chief Evangelist. So I’ve seen him present a few times and always good. This one was about the voice, chatbots, Lex, so AWS Lex which is their language parser, which is part of what Alexa uses. So he was saying that the voice interface is hugely powerful for providing access to the differently abled and disenfranchised people. So whether that’s someone who doesn’t have the ability to see, read, or whatever but that has the ability to speak and hear, obviously, would be pretty important for voice, that that actually is a connection to the internet which is basically becoming the world. So older people, so the people who are less able to use phones and computers, as long as they’ve still got their voice and hearing, then they haven’t been cut off from the rest of the world, and I remember hearing that someone who dealt with people at a retirement home or possibly a nursing home, that it actually was comforting because it was almost like company which is kind of sad to hear as a human but I could see myself being just as happy talking to a computer as to other people sometimes. So Lex, the parser, it works with more than just Alexa though, so it works well with Facebook Messenger and Twilio, and that’s just a few of the things, so it’s becoming an easy way to interface with all the learnings they have, both with text and with voice to your offering, connecting your application through some sort of conversational interface, whether that’s a chatbot or a voicebot. I think that that’s really interesting where that is headed, as long as it’s done right. I’m still unhappy with the results from Siri. Unfortunately, I can’t buy an Alexa easily here in Ireland. [29:39] could but they don’t ship here. They do ship, they just don’t ship those devices, so I’d have to buy a service that lets you ship it somewhere in the UK, then they bring it over here, just, I’m not that interested on getting one yet, just not enough time to play with it. Another funny thing is that Alexa gets told “I love you” a lot and I’m guessing that it’s more than any of us get told that, and I found out that AWS has a Twitch, twitch.tv/AWS. At first, I was like, what? What? Why? Why? And then my second thought was, hmm, I should check that out. I wonder what they’re showing there. So that was a pretty good demo, they created a chatbot with Lex which looks like all the other tools that are used to do that but it’s the power behind it that I’m sure is where the real usefulness lies.

And then the last one was Alan Kiernan of Ryanair and Cation Dynamics h I think I’m saying that right. I hope so. So Ryanair is an airline not usually known for anything other than cheap flights but supposedly, they’re trying really hard to reposition themselves as a customer service brand. There were rumors at one point that they were going to charge to use the toilets. I don’t think that ever actually happened and I don’t think - well, I don’t know whether or not it was really discussed, but there were rumors of it, probably just a joke, but they’ve created an Alexa skill that allows you to search for travel - flights and hotels, and a few other types of accommodation through Alexa and he was using one of the - I can’t remember the name of it but it’s basically a visual version of Alexa, so there’s a screen, looks kind of like a tablet or a bit like one of those electronic picture frames but what’s interesting to me about it is that that device reminds me of a really small TV, so Amazon is basically one small step away from an Amazon TV with picture-in-picture of Alexa. So you can tell your TV to change the channel and order stuff for you, and all the things Amazon wants you to do. That could be pretty interesting. I don’t know if I would want one because I still need to try Alexa or the dot but I’m excited about where voice technology is going. All of us older nerds love the old sci-fi, “Computer, do something for me.” That would be pretty nice if we could interface with the actual real world the way that you saw it in Star Trek, but that’s enough for today. Well, today, there’s another episode but that’s enough for this episode. So thanks for listening, talk to you later.

Until next time, remember, any sufficiently advanced technology is indistinguishable from magic.