Anthony Deighton: So the way to think about this is what would you need to do as a human being if you wanted to make the data better? So one of the things you might do is standardize a lot of the data. In fact, if you look at what a data engineer does when they do a master data management project, often times the upfront work is data standardization and harmonization. And so yeah, we do that. We’ll standardize all the addresses and phone numbers, email addresses. Why? Because those are really important signal for the model to be able to resolve the enemy.
You also start thinking about ideas like well duplication, sure, but also verifying that duplication. So if I give you three records and I give you various misspellings of Lydonia, right? And sometimes it’s Lydonia, Inc and sometimes the address is correct and sometimes it’s wrong. What do you do as a human being? You think to yourself, is there a real company in the world called Lydonia? Oh, there is? Great. Then likely these are all the same entity. In fact, they’re all likely to be this real world thing called Lydonia.
So that same concept favored us. So we will help resolve the entity by grounding it in ground truth. We call this a verified match. And part of the enrichment capabilities, we also want to produce a golden record.
What do I mean by that? If I give you five systems worth of data and we’re looking for the company name, which system do you trust or which third party do you trust? Maybe you say you know I always trust SAP for the company name but for the address I want to go to DMBB. I want to get the DMP address because that’s more valid. Great. You know we don’t have a strong opinion there. The system wants to be able to resolve that and deliver that back to you as a golden record.
So last thing I would point to is all of the real time APIs. It’s really important when you have a system like this, you’re thinking about mastery data, to think about how you tie it into your operational processes. And fundamentally that’s about opening these APIs and tying them into these systems. So when you on board a customer, that entity in the system as soon as that customer is onboarded so other systems or people can have that to resolve their case.
Kevin Scannell: We have one of our business partners that wants to give us a set of accounts to go out and cover for them for broader reach on their part and they try to get us that data and couldn’t and we found out they had a data problem. So we’re actually moving Tamr in there right now as we speak to go in and clean up their data so they can do this basic thing of just doing data fencing and giving us a set of accounts that we can cover on their behalf because they’re a pretty small salesforce.
Anthony Deighton: So look the question of who are my customers, where is my supply chain, where are my supply chain risks, how do I do cross sell and upsell. These are basic questions and many businesses struggle. And even if they can answer that question in a specific business unit or division or geography, being able to resolve that across many different systems against those dimensions like previous mentioned, those are hard things to do.
So you know this is a big problem but also a big opportunity. And we spent a lot of energy talking about the customer domain and to be clear it’s almost certainly the best place for most organizations to start. If you can’t understand your customers, there’s really no point in thinking about other domains. But there are many other domains, products, parts, locations, employees. Like you can start thinking about all the different kinds of data.
And really you know a simple way to think about Tamr is if I could have all the best data in my organization organized together in entities and available to everyone in my business through APIs, through analytical dashboards, in their operational systems, what would that mean for my business? Would that make me more efficient, more effective, serve my customers better?
Kevin Scannell: I would think so. Yeah, for sure. Anthony, this has been a great conversation. I appreciate you coming in today. It’s always good to see you, my friend.
Anthony Deighton: Pleasure.
