Interview with Dr. Thom Ives, Sr. Data Scientist at Echo Global Logistics

Last week, we got an amazing opportunity to interview Dr. Thom Ives, Sr. Data Scientist at Echo Global Logistics. With 30+ experience in the field of business and engineering, he is now a mentor and a data influencer for many growing minds in the data space.

In this interview, he shares with us his super interesting insights, inspiring and valuable experiences, and much more. We enjoyed every minute of this session and we hope you enjoy it too.?

Let’s get started! ??

Talking about challenges, as a data scientist what are the different problems you have faced when it comes to handling data?

But then there's the other side of the spectrum. Okay, we've got the tools. Now, what state is the data? That can be the bigger challenge because I've been in roles where you had to have special tribal knowledge to know which database to look at, and which fields you would need.?

Now, the nice thing is, that I have always been able to find tribal historians that could help. But I've even contemplated. How could I build my own tool to discover all the tables, discover all the columns, and start to make sense of their relations to one another?

You find your biggest pain points and you try to get the data clean and as many operations automated as soon as possible so that you can focus on the higher-level conceptual work.?

But I think someone that's fresh out of school, who's only been building their vertical knowledge faces problems when they come up against this tribal knowledge database which requires a lot of cleaning because there's a lot of dirty data that matters.?

It was a great, great answer and maybe I'll take the topic in a slightly different direction and emphasize something you said before, you see that the customer is the end stakeholder of data and thus, you in the data science department, have to keep the customer in the back of your head when working with data.?

In companies that are quite small, it was often an issue to really split tasks in the data team where they have one set of people who work on these longer projects, which really emphasized creating value, while the other set works on solving tickets from different customer-facing departments like operations. And that's why I wanted to ask how that looks like in the companies you have been working with so far?

I don't see data science work as one-dimensional and all about modeling, quite the contrary, when people come to me for mentoring. One thing I encourage them to do is get good at visualizing data, the better a data storyteller, the more helpful you'll find visualization towards your modeling, and that 80 plus percent we spend on getting the data ready for modeling. Really gives you 80% or more of the value you can get back to the organization.?

When you have your current data assets and when you have the business needs, don't just run to the top business needs and start working. We say, look at your current data assets, look at which project can move ahead, the quickest based on current data assets. With that also make the company aware you need more data assets here for us to get work done.

So speaking about that, How good are your current company and the many other companies you worked with before, in having that communication channel between the data department and other business units like sales?

Well, this is a great line of questioning and I was thinking back about the hard lessons. When I'm training people younger than me now to avoid my stupid mistakes. I sum it up in one word, care. Just care and I'd like to sum it up in another word - beer.?

I always believe that the best way to communicate data efforts or ideas is to talk it out directly face to face or by a video call rather than emailing or messaging the person. This is because significant business ideas or efforts need to be communicated in a proper channel to avoid unwanted interpretations and also to pass on the essence of what that actually is all about. So I always feel like giving a little more care to what you do and presenting it the right way.?

My engineering thinking and my data science thinking have always gone hand in hand with business. It's just that over time, I finally stopped sucking at the combination of the two.?

Let’s jump a little bit further into data enablement. What we currently see is that especially in startup business units are often the company's stakeholder or end stakeholder of data. So the data you prepare is either for the customer or if it's inside the company often for the operations or sales units, to just offer a really strong customer experience. However, they often do not have the ability to work with data themselves and have to go back to the data department for every single small thing.

And here we want to step in and help these departments to have access to data themselves and know, okay where to find it. We also provide a SQL tool that helps them to just create simple workflows, and simple automation in their daily routine to not have to go to the data department for every single new automation task because as you stated before, companies need to automate as many operations as possible and so we see our chance there. So what is your opinion on that?

There are these companies that can't afford a full-blown data science team or a full business intelligence team. And I think the role that you guys can play for companies like that is crucial because you can say, let us come in and meet your data automation and visualization needs. You're getting some great return on your data or maybe figuring out what Data, you could begin to collect better in organizing better so that you can get more data guidance, and then as needed we can help you build up your team too if that's what Datamin does and to me, that's a great space to be in this early periods of our obvious data age explosion.

I love the movie called Current Wars. It's the story of the interactions of George Westinghouse, Thomas Edison, and Nikolai Tesla, during the early age of electricity in our country. And I like to reference that because you see how there was some dirty, messy politics and stuff going on. We are kind of like that right now in the database age. We don't have a lot of clarity on terminology for both functions and roles and the techniques, but we're starting to understand each other better. We're starting to have more finely tuned terminology for various areas, but it's still messy and will be messy until things settle out.

The advantage to groups like yours is so much of what needs to be done in the early stages is the same as anywhere else. Why hire a team to build it from scratch when you guys have already been doing it and you can replicate tools you use for one client without revealing secrets to go, just basically, make a few configuration changes, and now they're benefiting from all this leveraging you're doing from effectively helping these other groups.?

So I think Datamin has to prove to their clients that working with them is more effective than building a team from scratch and building those tools from scratch. But also once you get to the next level, maintaining your relationship with Datamin is also more cost-effective than maintaining your integrity.

They don't want that mundane repetitive work. They want to operate as humans and let's face it at a certain point. What you guys are providing is super important. Against mundane and repetitive tasks - why have humans do that, when they can do higher-level functions?

Very true Thom, It was a great answer.

-------------------------------

Thank you very much Thom for taking the time and sharing super interesting insights with us and we loved our conversation with you.

We hope you enjoyed this article as much as we did ??

Well folks, if you are still curious and interested in knowing more about other different topics we covered in the interview session with Thom !!

Then tap on the link to our Medium article to unlock the full version of our interview session with Thom ? Interview with Dr. Thom Ives

-------------------------------

If you’re interested in Datamin’s product, reach out directly to our team, we are here for you! ??DATAMIN