Monday 13 November 2017

Podcast Interview with Nicolas Mervaillie, GraphAware

Here's another great interview with a long time Graphista that has done a lot of really interesting work in our French community, and is now having lots of graph-fun at GraphAware: Nicolas Mervaillie. Nicolas has been and still is working on some really cool stuff - so a chat was long overdue! Here's our recording - including a fancy new jingle from PremiumBeat ("Fantastic Voyage", by Olive Musique):

Here's the transcript of our conversation:
RVB: 00:00:03.275 Hello everyone, my name is Rik, Rik Van Bruggen from Neo and here I am again on a Skype call, recording the next episode in our Graphistania podcast. And today I have a-- I would say an oldtimer in our Neo4j community on the other side of this Skype call, all the way from Lille in France, and that's Nicolas Mervaillie. Hey Nicolas, how are you? 
NM: 00:00:26.083 Hey, good morning Rik. Thanks for inviting me.  

RVB: 00:00:28.790 Absolutely, yeah. It was something that I should have done probably a long time ago because you've been part of the French community for such a long time. But I finally got round to it, so I'm happy [laughter]. Nicolas, why don't you introduce yourself to our listeners. Who are you and what's your relationship to the wonderful world of graphs? 
NM: 00:00:50.515 Okay. I'm a consultant at GraphAware. We are a company dedicated to Neo4j and we help people succeed with their projects by doing consultancy and training and development around Neo4j, okay? And as part of GraphAware, I am doing some client projects, and also most of my time I am working on the Spring Data Neo4j and Neo4j OGM. 
RVB: 00:01:22.025 So, but you've been part of the Neo4j community for quite some time, right? Before you joined GraphAware
NM: 00:01:26.869 Yeah. I started maybe four or five years ago by doing a till receipt analysis project. So it's been quite a while. 
RVB: 00:01:39.772 I remember that. That was a fantastic project, right? You were analysing receipts from a point of sales, cashier thing, right? 
NM: 00:01:50.119 Yeah, that's right. It started when this company realised that the BI teams couldn't do it in the timeframe and on budget. So we started looking at it, and in a few hours we realised with the neo4j console, and we were quickly convinced that Neo4j could do the job, and that's it. A few days later, we were importing millions of receipts into Neo4j. 

RVB: 00:02:21.320 Fantastic. That sounds like a great use case and probably that's a good segue to kind of explain how and why you got into this world. What's your attraction to graphs? 
NM: 00:02:34.777 Well, it all started with this project, I mean, and after that, I got a lot in touch with the community around Neo4j, which is a very special thing to me, around Neo4j, because the community is fantastic. And after that, I always kept in touch with the guys in France, or at the occasional graph connects. And yeah, it's always such a pleasure to see and discuss with people, such special people. 
RVB: 00:03:08.968 Yeah, absolutely. And was there a particular reason why you thought that Neo4j was such a great fit for that ticket analysis project? Or for other projects specifically? Why do you think it was a good fit? 
NM: 00:03:22.687 Well, because it allowed us to model things quite like in the real world. We were amazed at how fast we got up and running with all this. It was really a matter of days. And yeah, this is really what I like. The fact that solving problems really quickly, and I think, yeah, Neo4j product [inaudible] are a great technology, to really go fast on finding solutions. 
RVB: 00:03:59.298 And you're someone who's quite technical, right? You come from the Java world. Is that right? 
NM: 00:04:03.363 Yeah, before discovering Neo4j, I'd been working since 15 years or 18 years doing consultancy with Java and Spring. In general, coaching teams, doing some technical leads or technical expertise around Java. And helping people develop better applications. 
RVB: 00:04:32.342 So tell us a little more about this work that you've been doing on the Spring Data Neo4j (note: there's an upconing online meetup about SDN that you may want to take a look at!), because that's been around for quite some time, but it had a big refresh, I think, not too long ago, right? You guys really redeveloped most of it, I think. 
NM: 00:04:47.549 Yeah. We just released version five, and we've been working on it for a bit less than one year. About Spring Data Neo4j, it's a framework that makes developing Java or JVM-based application faster. There's quite a lot of big companies using it, and with these new versions we did a lot of new features, like making the programming model more fluent, adding powerful features, and be sure that this was all ready for enterprise use cases, like really good cluster support, or working on going toward reactive programming models or event-based architectures. 
RVB: 00:05:48.415 Cool. And as I understand it - you have to forgive me here because I am not a Java programmer - but as I understand it, part of Spring Data Neo4j is the object-to-graph mapper. Right? Is that true? 
NM: 00:06:05.056 Yes, that's right. 
RVB: 00:06:07.639 And how does that work? Because people might know object-relational mappers, but object-to-graph mappers are quite interesting in my book, because they typically tend to be a lot lighter than the object-relational mappers. Right? 
NM: 00:06:26.053 Yes, because they are dedicated to Neo4j in this case. But we can really see them as equivalent. Yeah, OGM, the Neo4j OGM, is really what's under the hood of Spring Data Neo4j. It's really the engine behind Spring Data. Spring Data on Neo4j is what makes integrating this OGM into the Spring ecosystem, and making it fluent for people and enterprises that often use Spring Framework when they get into larger projects, to make it really fluent and make the user of Neo4j very easy on projects that are already using Spring. 
RVB: 00:07:17.874 But as I understand it, you can use the object-to-graph mapper without Spring. Right? That's possible. Right? 
NM: 00:07:23.576 So when you're not a Spring shop, using just plain OGM makes it available for plain Java applications to use Neo4j. That's right. 
RVB: 00:07:33.524 Well, what I think we'll do is we'll include a couple of links to all of this beautiful stuff in the transcription of the podcast, and then we can just point people in the right direction. But I'd like to talk a little more about the future if you don't mind. What does the future hold for you, Nicolas, for GraphAware, and obviously also for our industry? How do you look at that? You've been in the ecosystem for quite some time so you probably have a good perspective. 
NM: 00:08:06.761 Well, a lot of things come to mind. I hope that the future will show that Neo4j can help make a better world. I mean when we see the Panama or Paradise Papers more recently, I find it really great-- 
RVB: 00:08:23.502 Oh, wasn't that great? 
NM: 00:08:23.939 Yeah. 
RVB: 00:08:25.352 That was fantastic. 
NM: 00:08:26.890 To see such amazing things coming, with the help of Neo4j and Linkurious and GraphAware on other projects like that. Otherwise, I mean I think in the future we'll see graph databases more and more integrated into solutions. In the past, for example, I've been working on project information management system, and I see how it can be painful with traditional databases and how slow and complicated it can be to use traditional technologies to implement such solutions, and I think graph databases are a real game changer for solutions like that. So I expect to see more and more use of graph databases. I'm also amazed to see all the things we can build on top of Neo4j. For example, GraphAware, we're working a lot on natural long wedge processing and building knowledge graphs with Neo4j. And it's pretty much work in progress, yet. But... 
RVB: 00:09:36.027 Yeah. I saw that at GraphConnect. Christophe was demonstrating that. Right? Christophe Willemsen? 
NM: 00:09:39.372 Yeah, and building chatbots around it. And I'm really impatient to see where on this will lead us. 
RVB: 00:09:48.723 But as I understand it, there's something else in your future. I hear that you're working on a book. Right? 
NM: 00:09:53.980 Yeah, that's right. A French book which is going to be published this month. So I'm a modest coauthor of this book. The main man behind this is Sylvain Roussy which is the Lyon meetup organiser -- 
RVB: 00:10:12.147 He wrote another book. Right? 
NM: 00:10:14.001 Yeah, it was a first version of the book a few years ago, and there's Nicolas Rouyer who organizes the meet-ups in Toulouse, of Neo4j. And it's a quite different book from the usual you can read around. I mean it's a real story, with characters talking and discussing how to build a complete solution using Neo4j. So it really starts with the design and the needs for this solution, and goes all through until they put it in production and monitor it and backups. That's all. 
RVB: 00:10:59.051 Sounds really interesting, and think we need something like that in English as well actually. So I'm looking forward to reading it. Absolutely. So Nicolas, I think we're going to wrap up the podcast. As you know, we keep this quite short so that people can enjoy it on their commutes and stuff like that. But I want to thank you for coming online and doing this with me. It was a real pleasure and I look forward to meeting you at one of the future community events. Right? 
NM: 00:11:29.093 Yeah. Thank you, Rik. Thank you very much. And yeah, see you at the next event. 
RVB: 00:11:33.948 Absolutely. Talk to you soon. Bye-bye. 
NM: 00:11:36.056 Bye-bye.
Subscribing to the podcast is easy: just add the rss feed or add us in iTunes! Hope you'll enjoy it!

All the best


No comments:

Post a Comment