Experiencing Data w/ Brian T. O’Neill

Today we are joined by the analytics “man of steel,” Steve Bartos, the Manager of the Predictive Analytics team in the steel processing division at Worthington Industries. 😉 At Worthington, Steve is tasked with strategically driving impactful analytics wider and deeper across the division and, as part of this effort, helps ensure an educated, supported, and connected analytics community. In addition, Steve also serves as a co-leader of the Columbus Tableau User Group.

On today’s episode, Steve and I discuss how analytics are providing internal process improvements at Worthington. We also cover the challenges Steve faces designing effective data-rich products, the importance of the “last mile,” and how his PhD in science education shapes his work in predictive analytics.

In addition, we also talk about:

• Internal tools that Steve has developed and how they help Worthington Industries.

Twitter: @OlDirtyBarGraph

Steve Bartos LinkedIn

“Seeing the way analytics can help facilitate better decision making, doesn't necessarily come with showing someone every single question they can possibly answer, waiting for them to applaud how much time and how much energy and effort you'd saved them.” - Steve Bartos

“It's hard to talk about the influence of different machine parameters on quality if every operator is setting it up based on their own tribal knowledge of how it runs best.” - Steve Bartos

“I think bringing the question back to the user much more frequently, much sooner and at a much more focused scale has paid dividends”. - Steve Bartos

“It's getting the people that are actually going to sit and use these interfaces involved in the creation process… they should be helping you define the goals and the problems… by getting them involved, it makes the adoption process a lot easier.” - Brian O’Neill

“It's real easy to throw up some work that you've done in Tableau around a question that a manager or an executive had. It's real easy to do that. It's really difficult to do that well and have some control of the conversation, being able to say, here's what we did, here was the question, here's the day we use, here's how we analyze it and here's a suggestion where making and now let's talk about why and do that in a way that doesn't lead to an in-the-weeds session and frustration.” - Steve Bartos

Paul Mattal is the Director of Network Systems at Akamai, one of the largest content delivery networks in the U.S. Akamai is a major part of the backbone of the internet and on today’s episode, Paul is going to talk about the massive amount of telemetry that comes into Akamai and the various decision support tools his group is in charge of providing to internal customers. On top of the analytics aspect of our chat, we also discussed how Paul is approaching his team’s work being relatively new at Akamai.

Additionally, we covered:

• How does Paul access and use internal customer knowledge to improve the quality of applications they make?

Akamai

Twitter @pjmattal

Paul Mattal on LinkedIn

Paul Mattal on Facebook

“I would say we have a lot of engagement with [customers] here. People jump to answering questions with data and they’re quick. They know how to do that and they have very good ideas about how to make sure that the approaches they take are backed by data and backed by evidence.” — Paul Mattal

“There’s actually a very mature culture here at Akamai of helping each other. Not necessarily taking on an enormous project if you don’t have the time for it, but opening your door and helping somebody solve a problem, if you have expertise that can help them.” — Paul Mattal

“I’m always curious about feedback cycles because there’s a lot of places that they start with telemetry and data, then they put technology on top of it, they build a bunch of software, and look at releases and outputs as the final part. It’s actually not. It’s the outcomes that come from the stuff we built that matter. If you don’t know what outcomes those look like, then you don’t know if you actually created anything meaningful.” — Brian O’Neill

“We’ve talked a little bit about the MVP approach, which is about doing that minimal amount of work, which may or may not be working code, but you did a minimum amount of stuff to figure out whether or not it’s meeting a need that your customer has. You’re going through some type of observation process to fuel the first thing, asset or output that you create. It’s fueled by some kind of observation or research upfront so that when you go up to bat and take a swing with something real, there’s a better chance of at least a base hit.” — Brian O’Neill

“Pretend to be the new guy for as long as you can. Go ask [about their needs/challenges] again and get to really understand what that person [customer] is experiencing, because I know you’re going to able to meet the need much better.” — Paul Mattal

Dr. Andrey Sharapov is a senior data scientist and machine learning engineer at Lidl. He is currently working on various projects related to machine learning and data product development including analytical planning tools that help with business issues such as stocking and purchasing. Previously, he spent 2 years at Xaxis and he led data science initiatives and developed tools for customer analytics at TeamViewer. Andrey and I met at a Predicitve Analytics World conference we were both speaking at, and I found out he is very interested in “explainable AI,” an aspect of user experience that I think is worth talking about and so that’s what today’s episode will focus on.

In our chat, we covered:

• Lidl’s planning tool for their operational teams and what it predicts.

Also, here’s a little post-episode thought from a design perspective:

I know there are counter-vailing opinions that state that explainability of models is “over-hyped.” One popular rationalization uses examples such as how certain professions (e.g. medical practitioners) make decisions all the time that cannot be fully explained, yet people believe the decision making without necessarily expecting it to be fully explained. The reality is that while not every model or end UX necessarily needs explainability, I think there are human factors that can be satisfied by providing explainability such as building customer trust more rapidly, or helping convince customers/users why/how a new technology solution may be better than “the old way” of doing things. This is not a blanket recommendation to “always include explainability” in your service/app/UI; I think many factors come into play and as with any design choice, I think you should let your customer/user feedback help you decide whether your service needs explainability to be valuable, useful, and engaging.

Andrey Sharapov on LinkedIn

Explainable AI- XAI Group (LinkedIn)

“I hear frequently there can be a tendency in the data science community to want to do excellent data science work and not necessarily do excellent business work. I also hear how some data scientists may think, ‘explainable AI is not going to improve the model’ or ‘help me get published’ –  so maybe that’s responsible for why [explainable AI] is not as widely in use.” – Brian O’Neill

“When you go and talk to an operational person, who has in mind a certain number of basic rules, say three, five, or six rules [they use] when doing planning, and then when you come to him with a machine learning model, something that is let’s say, ‘black box,’ and then you tell him ‘okay, just trust my prediction,’ then in most of the cases, it just simply doesn’t work. They don’t trust it. But the moment when you come with an explanation for every single prediction your model does, you are increasing your chances of a mutual conversation between this responsible person and the model…” –  Andrey Sharapov

“We actually do a lot of traveling these days, going to Bulgaria, going to Poland, Hungry, every country, we try to talk to these people [our users] directly. [We] try to get the requirements directly from them and then show the results back to them…” –  Andrey Sharapov

“The sole purpose of the tool we built was to make their work more efficient, in a sense that they could not only produce better results in terms of accuracy, but they could also learn about the market themselves because we created a plot for elasticity curves. They could play with the price and see if they made the price too high, too low, and how much the order quantity would change.” –  Andrey Sharapov

My guest today is Gadi Oren, the VP of Product for LogicMonitor. Gadi is responsible for the company’s strategic vision and product initiatives. Previously, Gadi was the CEO and Co-Founder of ITculate, where he was responsible for developing world-class technology and product that created contextual monitoring by discovering and leveraging application topology. Gadi previously served as the CTO and Co-founder of Cloudscope and he has a management degree from Sloan MIT.

Today we are going to talk with Gadi about analytics in the context of monitoring applications. This was a fun chat as Gadi and I have both worked on several applications in this space, and it was great to hear how Gadi is habitually integrating customers into his product development process. You’re also going to hear Gadi’s interesting way of framing declarative analytics as casting “opinions,” which I thought was really interesting from a UX standpoint. We also discussed:

• How to define what is “normal” for an environment being monitored and when to be concerned about variations.

LogicMonitor

“The barrier of replacing software goes down. Bad software will go out and better software will come in. If it’s easier to use, you will actually win in the marketplace because of that. It’s not a secondary aspect.” – Gadi Oren

“…ultimately, [not talking to customers] is going to take you away from understanding what’s going on and you’ll be operating on interpolating from information you know instead of listening to the customer.” – Gadi Oren

“Providing the data or the evidence for the conclusion is a way not to black box everything. You’re providing the human with the relevant analysis and evidence that went into the conclusion and hope if that was modeled on their behavior, then you’re modeling the system around what they would have done. You’re basically just replacing human work with computer work.” — Brian O’Neill

“What I found in my career and experience with clients is that sometimes if they can’t get it perfect, they’re worried about doing anything at all. I like this idea of [software analytics] casting an opinion.” — Brian O’Neill

“LogicMonitor’s mission is to provide a monitoring solution that just works, that’s simple enough to just go in, install it quickly, and get coverage on everything you need so that you as a company can focus on what you really care about, which is your business.” — Gadi Oren

My guest today is Carl Hoffman, the CEO of Basis Technology, and a specialist in text analytics. Carl founded Basis Technology in 1995, and in 1999, the company shipped its first products for website internationalization, enabling Lycos and Google to become the first search engines capable of cataloging the web in both Asian and European languages. In 2003, the company shipped its first Arabic analyzer and began development of a comprehensive text analytics platform. Today, Basis Technology is recognized as the leading provider of components for information retrieval, entity extraction, and entity resolution in many languages. Carl has been directly involved with the company’s activities in support of U.S. national security missions and works closely with analysts in the U.S. intelligence community.

Many of you work all day in the world of analytics: numbers, charts, metrics, data visualization, etc. But, today we’re going to talk about one of the other ingredients in designing good data products: text! As an amateur polyglot myself (I speak decent Portuguese, Spanish, and am attempting to learn Polish), I really enjoyed this discussion with Carl. If you are interested in languages, text analytics, search interfaces, entity resolution, and are curious to learn what any of this has to do with offline events such as the Boston Marathon Bombing, you’re going to enjoy my chat with Carl. We covered:

• How text analytics software is used by Border patrol agencies and its limitations.

• Basis Technology

“One of the practices that I’ve always liked is actually getting people that aren’t like you, that don’t think like you, in order to intentionally tease out what you don’t know. You know that you’re not going to look at the problem the same way they do…” — Brian O’Neill

“Bias is incredibly important in any system that tries to respond to human behavior. We have our own innate cultural biases that we’re sometimes not even aware of. As you [Brian] point out, it’s impossible to separate human language from the underlying culture and, in some cases, geography and the lifestyle of the people who speak that language…” — Carl Hoffman

“What I can tell you is that context and nuance are equally important in both spoken and written human communication…Capturing all of the context means that you can do a much better job of the analytics.” — Carl Hoffman

“It’s sad when you have these gaps like what happened in this border crossing case where a name spelling is responsible for not flagging down [the right] people. I mean, we put people on the moon and we get something like a name spelling [entity resolution] wrong. It’s shocking in a way.” — Brian O’Neill

“We live in a world which is constantly shades of gray and the challenge is getting as close to yes or no as we can.”– Carl Hoffman