Thursday 6 October 2016

Podcast Interview with Alessio De Angelis, Whitehall Reply

I have said it before and I will say it again, but the joy of this podcast - for me, at least - is the fact that I get to talk to all of these wonderful people out there. Today's conversation is with a really cool and talented young software developer out in Rome, Italy, who has had a passion for graphs for a while - and is now applying it to Neo4j and NOSQL database approaches. Here's my chat with Alessio De Angelis:

Here's the transcript of our conversation:
RVB: 00:02.287 Hello everyone. My name is Rik, Rik van Bruggen from Neo Technology, and here I am recording another Graphistania Neo4j podcast, and today I'm joined by someone from a place with a lot better weather than where I am, that's for sure. Alessio De Angelis from Rome in Italy. Hi Alessio. 
ADA: 00:21.861 Hi Rik. 
RVB: 00:23.151 Hey, it's great to have you on the podcast. Thank you for making the time. 
ADA: 00:26.929 Thank you for inviting me. 
RVB: 00:28.415 Very good. Alessio, our listeners probably don't know you yet, so I read some of your work on the GraphGist Channel Challenge (Santa's shortest weighted path), but you may want to introduce yourself. Who are you, what do you do? 
ADA: 00:44.241 Yes, sure. I'm now working as an IT consultant, in particular involve into big data, data warehouse, and SQL projects, but my love with the graphs started a long time ago, probably three years ago, because at my university I was working on a Master's thesis, where I had to do a recommendation engine, in the cultural heritage domain, and I was looking for a tool that was helping me to profile the user interest, storing the places, historical monuments he and his friends visited according to their social network accounts, together with the extra information delivered by linking the data  in graphs. And Neo4j was the perfect database to achieve all those tasks. So, I really fell in love with this database.
RVB: 01:56.472 That's great to hear. Fantastic. I mean, there's a lot of people that are actually using Neo4j in the cultural heritage space. I've actually interviewed a couple of them on this podcast already, so I'll send you some links and I'll put them on the transcription as well (note: see posts like this one, interviewing Iian Neill about his Codex, or this one, interviewing Lorenzo Speranzoni about his work on Van Gogh's journey). So, that was a really great use-case, and then you also decided that you wanted to do something for Santa Claus as well? 
ADA: 02:21.394 Yeah [laughter]. 
RVB: 02:24.278 What was that about? 
ADA: 02:26.193 I started looking for-- I was in the Neo4j main page-- actually, no, in the Twitter account, and I read up about these GraphGist Challenge, and say, "Whoa, I want to participate." And then I was seeing the domains, and I saw Santa Claus. Nice. Let's think about Santa Claus life. And since I was a little kid, I was wondering how could Santa Claus manage to reach every children in the world, and give the presents before they fall asleep. 
RVB: 03:11.683 Because he has a graph database. Now we know. 
ADA: 03:13.636 Yes [laughter]. 
RVB: 03:18.643 Very funny. 
ADA: 03:18.854 He managed to think about an algorithm for searching the shortest path in the graph of children all over the world. I think for sure that is the reason why Santa Claus managed to do it now in 2016. 
RVB: 03:39.121 There is no doubt, I know he uses Neo4j [chuckles], but maybe he's actually using it with some of the new awesome procedures, the APOC, because I don't know if you are familiar with that but the weighted shortest path to calculations, you can do those with algorithms like Dijkstra and A* [crosstalk]. 
ADA: 03:59.679 Yes. 
RVB: 04:01.044 And in the APOC, in the awesome procedures that we have now in Neo4j 3.X, there's those algorithms and you can call them from Cypher. I think next Christmas you have to go to Santa Claus, and you have to tell him how to improve his database. 
ADA: 04:16.610 Yes. For sure [laughter]. 
RVB: 04:19.417 Very cool. So, why is it so attractive to you? Why did you fall in love you think with the wonderful world of graphs, is it the model, is it the performance, what is it about for you? 
ADA: 04:33.153 Firstly, because as the Neo4j slogan says, graphs are everywhere. Probably we don't think about it but really graphs are everywhere, and probably you can model all domain in a graph way. And then in particular, I was attracted by Cypher, because it's so short syntax, but really powerful. I really love the pattern matching, and Santa Claus in his team managed to do all the search of the shortest way weighted path with just Cypher and his main constructs like-- 
RVB: 05:22.258 Fantastic. That's a great summary. I mean if Santa Claus can use it then everyone can, right? 
ADA: 05:27.348 Yes. 
RVB: 05:28.332 It's obvious [laughter]. And, is there anything that you think that would really make it even more awesome, that you think we should really be adding to Cypher, or Neo4j, or maybe looking a little bit into the future? What do you think about that? What do you think the future holds for this industry? 
ADA: 05:53.424 I think it's really growing because, even in Italy, much more companies are getting through this database, and if I think several years ago no one was knowing that database is probably just at university. In some courses, I was listening to people talking about graph databases, but now I think it's really a reality. In the future, it would be a really-- much more projects would be involved no SQL databases. 
RVB: 06:38.832 Is there a particular kind of project where you think that it would really fit best, or what do you think would be the sweet spot for the graphs in the next couple of years? 
ADA: 06:48.486 Yeah. For sure, in our research, operations, course, we did so many , like for searching the best solution, or-- yeah. So, I think each one of these theoretical aggregates can have an improvement if it's applied with a graph database, like Santa Claus did with just a normal shortest weighted path search algorithm, improved by Cypher language. 
RVB: 07:30.011 So, basically what you're saying is that there's a lot of just regular database applications that could benefit from it, and offer new functionalities, new [crosstalk]. 
ADA: 07:40.226 Yes, probably just some-- a algorithm that work on data as a graph, but just not considering the data as a graph itself. So, they try to travel among these data, and try to connect them, but without using a proper graph database. If they use it, I'm sure that their performance will improve. 
RVB: 08:12.035 I agree. I couldn't agree more, actually. Thank you so much for sharing that, Alessio, I really appreciate it. And, again, maybe next Christmas we'll have a new and improved [chuckles] engine for the pathfinding of Santa Claus. In the meanwhile, I think we'll try to keep these podcasts short and snappy, so I'm going to thank you for coming online, and sharing your experience with us. It's been a great time talking to you, and I hope to see you at one of the future Neo4j community or conference events, you know? 
ADA: 08:53.743 Yes. 
RVB: 08:55.047 That'll be great. Thank you [crosstalk]. Have a nice day. 
ADA: 08:58.099 Yeah, you too. Thank you for inviting me. 
RVB: 08:59.814 Bye. 
ADA: 09:00.891 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