Distributed Data Show Episode 66
Amanda sits down with Dr. Denise Gosnell at Graph Day to discuss highlights from her keynote.
0:15 Amanda: Welcome, and Graph Day is here!
0:37 Denise: Happy to be here and happy to be at graph day. One of favorite events. More like a family event for the graph community. So it's like a family reunions.
Amanda “laughs” at the thought of loving a family reunion.
0:50 Amanda: This is my first graph day!
0:55 Denise: Welcome to the family!”
1:15 Amanda: The 6 lessons of putting Graph into production. Really interesting connection to the typical journey of going through the journey of SimCity2000
1:33 Amanda: But Why SimCity?
1:34 Denise: Loved playing the game growing up. Loved being able to build something up from nothing on an empty playing field. And started playing the game again since it is available on VM. Found the 6 connections to how to be successful in the game and how to optimize to get your graph to work in production.
2:48 Amanda: I like how you said “an empty playing field”. Just a typical engineer and scientist way of thinking. We love that blank slate and we are going to make something.
3:00 Amanda: Really connected with your second point about “Knowing the Rules of the Game” this is an interesting point for living life, but how do you connect that to the world of graph?
3:20 Denise: It's easy to figure out how to win in SimCity. And in the graph community we have forgotten some of the rules of how to “win”. Easy to forget since we are flipping upside down how we think about data. Relationships first. How to measure the success of a graph in production really comes down to understanding your graph data and how many edges you have between two vertices.
4:20 Denise: If you don’t understand the number of relationships in your data, and what your trying to process you won’t understand how long your travesals will take.
4:32 Denise: Branching factor, if you do this blindly this is a non-linear operations under the hood
4:57 Denise: “Know THY DATA! Understand what your trying to do”
5:20 Amanda: This is a new trend, that we keep hearing. Know thy data, have domain experts, or become a domain expert.
6:00 Denise: Not always making the best choices about data modeling and look forward to learning more from others at this Graph Day and other Graph Days.
6:38 Amanda: You got a lot of questions at the end of your keynote. Many questions about graph databases about the ecosystem we are in. We have complicated answers to solve, so it's going to have complicated solution.
7:12 Denise: Need to remember 2 things. 1. Graph problems are much bigger than graph databases. You can think of this at a conceptional model but now we are starting to have tools to store this data. What other tools do we have to help solve these issues? You need a multimodal approach. 2. Remember why we joined this space? We have scars from what has worked and didn’t work. Remember the power of looking at relationships that will give new business value that we didn't have before.
9:13 Amanda Closes it out!