Thanks to another round of funding, Neo4j is investing in how its graph database technology can combine with artificial intelligence. Speaking to ZDNet from November's WebSummit in Lisbon, the company's founder, Emil Eifrem, discussed his plans.
ZDNet: Your latest round of funding raised $89m, how to you intend spending it?
Eifrem: I think there are two broad vectors that I want to invest in. The first is the increasing traction in the enterprise space. Over 80 per cent of the Fortune 100 are now using Neo4j. Twenty of the twenty-five biggest banks are using it along with the five biggest telcos. It's pretty deep in terms of adoption among the largest companies.
There [are] 200 companies here in Lisbon that we have great use cases for and there's no reason why a Portuguese bank doesn't have the same needs as a bank in London, so the need is there.
Obviously, to sell it to a company in Lisbon you want a mobile person, you want sales people, marketing collateral in Portuguese, support people in Portugal, all that good stuff. So, just scaling up operationally to meet that demand, that's the first vector.
The second vector is product. Investing in building out the graph platform is huge. But the biggest driver is the emergence of AI use cases. We have applications like eBay ShopBot and many others, which gives a voice interface to eBay, so shoppers can talk to the app and say, "Hey Google, I want to buy something on eBay" and eBay would then ask you what you wanted to buy and eBay would know your preferences so it would tailor its offer to you and then ask questions. What size? What colour? And so on.
That's one form of AI that is emerging but there are several more that we are doing, and we aim to really invest in the product roadmap and make sure that we can make the most of it.
You see AI as being very big for you?
Yes, it's a real driver of graph data.
It appears to be an intersection based on what a graph database is capable of when combined with AI?
There is something very intimate about the relationship between graph and AI.
Even a year ago, I could talk at a very high level about some of it but now we are seeing the really concrete use cases in production with customers.
SEE: How to implement AI and machine learning (ZDNet special report) | Download the report as a PDF (TechRepublic)
Now we are seeing the AI use cases -- be it knowledge graph, data extraction, graph-accelerated machine learning, AI flexibility, and those kinds of things. And I think, fundamentally, what we do is very horizontal.
It may start off with some early use cases where the pain is big enough and the value is big enough to make it worth it. But like Oracle in the '80s, it's not application or use case driven, it's fundamentally data driven.
One of the areas that you are getting into is the health sector. Can you show use cases there?
One of those that you have talked to is Germany's DZD centre which is looking at graph's use in diabetes research and I think it helps show that graph is a very good, generalised thing.
I've always been very hopeful that we are going to see a lot of usage in health care and life sciences based on the observation that if you look inside our bodies, there's a lot of graph in there.
Nature has a lot of graph -- everything from protein interactions, protein cells, and protein networks to truly ourselves. My favourite thing to talk about is how the human brain is a graph -- neuron connects itself to neuron. Clearly there's a lot of graph data inside the human body. That gives us a value for our technology inside that space. That's one angle. The other angle, is what I see all over the place, they are trapped by silos.
It's a bit like... you go back hundreds of years and the advancement of medicine as a field of science was based on observation -- trying out core science and biology. And then, all of a sudden, we found out this digital thing and what the scientists -- like DZD -- tell me is, we have found out all these things from silos and while there may be some incremental things to be found, by and large, the over-riding shift is being able to correlate across these silos.
AI software in action
This is what AI looks like (as sketched by AI)
IBM has some up with an AI image produced by an AI system tasked with coming up with an image of itself to accompany a New York Times series about the technology.
AI Edge X: The first 4G Industrial Gateway for AI on the Edge
Bridging the gap between the edge and the cloud in order to bring accelerated vision features and AI performance to the edge of the Internet of Things.
Nautilus launches AI platform to coach, train Bowflex customers
Max Intelligence is a cloud-based subscription service delivered through Samsung's Knox platform and tablets to move users toward health and fitness goals.
KT develops AI to detect telecom failures
The AI solution, named Dr Lauren, collects and analyses data produced by the network to find any fundamental flaws that cause telecommunication failures.
How Amazon's DeepLens seeks to rewire the old web with new AI
In a world where users interact with cloud-based servers using not only audio but video, text may already be out of fashion which may be fine with Amazon, which is testing a deep learning-enabled camera like it's going for broke.
LG CNS, Lunit to apply AI, cloud into X-Ray analysis
LG CNS will collaborate with medical AI start-up Lunit to use the cloud to analyze X-Ray videos within 15 seconds to diagnose lung diseases.