A view into AI in the context of integration.
Sramana Mitra: Let’s start by introducing our audience to yourself and SnapLogic.
James Markarian: I’m the CTO at SnapLogic. I’ve had a very unusual career in the Valley. In about 30 years, I’ve only had three jobs. I was at Oracle for a number of years in the earliest days. I was CTO at Informatica for about 15 years. I’ve been at SnapLogic for a little bit over a year now. I don’t talk about it too much, but I was at a venture capital company for about a year in between Informatica and SnapLogic.
Sramana Mitra: You go back with Gaurav for a long time?
James Markarian: I’ve known Gaurav coming up on 20 years now. It’s been a long relationship. I was hired into Informatica by somebody else that you know – Diaz Nesamoney.
Sramana Mitra: Let’s also talk a little bit about SnapLogic. Gaurav was here on several occasions. The audience is somewhat familiar with SnapLogic, but let’s put a little bit of an introductory context especially in the context of AI. SnapLogic is a provider of integration technology.
James Markarian: We cover what some folks might recognize as the gamut from what we would call application integration, which is process and message-oriented integration between applications. We also cover data integration, which is the integration between operational data stores and applications into analytics stores. A lot of problems don’t necessarily lend themselves to one thing or another, but they’re actually hybrids of those capabilities.
Sramana Mitra: Talk to me about what you are doing with AI. What trends are you seeing in AI and how does that pertain to your work?
James Markarian: I’ll give you a slight reference back to my Khosla Ventures days. I used to joke there that we should make entrepreneurs fill out a form when they walk in the door stating which technology area they’re disrupting using AI or machine learning. If you couldn’t tell us simply what area you were disrupting, then we didn’t want to talk to you. It was a little bit tongue-in-cheek. There was a belief that started for me then which continues now.
Every area of technology is currently or will be disrupted using AI and machine learning. That continues over to application and data integration. The first thing you would say is there’s a fundamental observation. The area of app and data integration has been addressed by packaged software versus hand-coding. Over the last 20 years, this has gone through that transition. My observation from my experience at the previous company and now is that, more or less, the process that you go through is still largely manual even though software is making it bitterer.
When you look at the steps that folks go through in understanding their data, it’s has always felt like we’re leaving something on the table. Generally speaking, we’re very good at training people to work the way that technology wants them to work. We’ve been very bad at making the technology work for people. It’s especially stark now in what I consider to be a glut of various things. A glut of hardware out there meaning cheap and easy to procure via Infrastructure-as-a-Service.
Also this glut in the form of interesting technology like AI and machine learning. It seems like the glut of hardware, the availability of this technology, folks that are skilled in it, and problems presented by integration seem like a natural fit to me. That’s what we’re working on exploiting nowadays.