25 Nov 2016 by Miro Marchi
Last month, the 5th edition of GraphConnect San Francisco took place at the Hyatt Regency SF. It was the biggest graph technology event ever and GraphAware proudly contributed as a sponsor, with one main talk, two lightning talks and our GraphHero stand ｡^‿^｡ This edition’s big announcement was the upcoming new landmark release of Neo4j 3.1, “The database for the connected enterprise”, which introduces a new state-of-the-art clustering architecture and new security architecture to meet enterprise requirements for scale and security. There will be a lot to say about this release, but you can already try the beta release as we have done!
It was a long and amazing week for us and for the entire Neo4j community. We started on Tuesday with our team meeting. GraphAware is a global company, spread among many countries and continents. GraphConnect events are great moments to gather almost all the team together, meet with our graphista friends from all around the world, and have fun! Anchor Brewing was the perfect place to hang out late, with local craft beers and the chef tacos.
On Wednesday we attended the Neo4j Ecosystem Summit meeting for deep and strategic discussions with Neo Technology and other Neo4j partner companies. This is a very important moment where the main players in the Neo4j community gather together for sharing important insights on graph technology on which we thrive. And we take good care of it!
The evening before the conference is the playground moment, where aspiring and veteran graphistas gather to play with the technology. The GraphHack theme of GCSF 2016 was data journalism, and dataset used spanned from ICIJ offshore data to elections, government opendata, and NASA data. 15 teams “competed for the luxurious prizes”, among which the incredible star-wars drone that our chief scientist was dreaming of (of course, he says it’s for his son)!
Our “dream team” with Alessandro and Christophe chose a dataset with US Senate press releases from the “Deep and interesting datasets for computational journalists” list by Dan Nguyen. Press releases contain text and it is a perfect dataset to try out GraphAware new Natural Language Processing (NLP) plugin, which uses Neo4j, Elasticsearch and Apache Spark to create a graph of sentences and tags, compute similarity between texts, perform sentiment analysis and other cool things you can do combining NLP with graphs, such as enriching the data with semantic ontologies.
For those of you that have never attended a GraphConnect conference, we can say it is the place to be if you want to find out more about graph databases and Neo4j in particular. Whether you are an expert or you just heard of this technology and all the complex use cases where it makes the difference, this is mission control.
You can find experts in graph visualization, graph database hosting, graph consultancy and software development, and organizations that happily implemented a Neo4j solution in their businesses. We and the other members of this connected and active community-of-practice, share our experiences using graph databases in different domains. We discuss successes and warnings, and everyone can learn from our talks, lightning talks, case studies, ‘how-to’ sessions, asking us directly at our stands or simply chatting during the fika coffee breaks or lunch times. The contents are organized by the Neo technology editors into different learning tracks, and targets a diverse audience, including IT executives, developers, and architects. Read more in Neo4j blog.
GraphAware participated deeply in this knowledge sharing and production: Luanne, our Principal Consultant, gave a talk about “Neo4j as a Key Player in Human Capital Management”, showing the hidden value in the relationships between people, tasks, skills, certifications, posts, experts… and how to make the most out of them with Neo4j.
Alessandro, our Chief Scientist, gave a lightning talk about “Taming text with Neo4j: The Graphaware NLP Framework”, explaining our new plugin.
(image (c) Neo Technology)
Christophe spoke on, “Cypher: Write Fast and Furious”, teaching developers how to increase performances of their graph queries. Two of our customers gave talks about their case studies: “Streamlining Processes with Neo4j at Glidewell Laboratories”, by Galit Gontar and Robert Edwards; and “Improving Service Quality for Cablevisión Customers with Network Topology Impact Analysis”, by Hernán Pérez Masci.
If some of you think it’s a pity having missed this and the other talks, don’t worry. The first videos are already being published at graphconnect.com, and there will be others to follow (and slides too). Stay in touch!
All the graph thinking and talking during the conference hours requires quality down time later, which is why GraphConnect culminated with a disConnect party. Buffet and drinks at a restaurant terrace on the Bay was the prescription, and Crystal Jade filled it perfectly.
Neo4j training filled the day on Friday. Five different courses were taught, ranging from introduction to Neo4j up to advanced deployment with Neo4j. We were there at our stand, where, by the way, our GraphHero and our special GraphAware hand-made cookies made many graphistas very happy!!! These cannot be enjoyed online unfortunately, so we recommend coming next year to sample for yourself all we have on offer!
(image (c) Neo Technology)
We had an Italian team dinner that evening, closing a week which was perfectly organized by Kateřina, our Operations Manager. It was great to reaffirm our relations. We are quite a family at GraphAware, and now we know we also have at least one great musician among us: Jasper! Not only is he our Senior Consultant, but he sang and played stunning blues grooves with his harmonica! We might go for a GraphBand next year!