Dean Robinson: The Data Archaeologist

Centric Consulting Data Architect Dean Robinson brings an archaeologist’s mindset to business data and business intelligence, using data artifacts to uncover hidden patterns and insights. Now, he uses AI to transform complex analyses into actionable stories and better decision-making. 

As a student, Centric Consulting’s Dean Robinson wanted to be an archaeologist. 

He imagined himself digging for artifacts, classifying them, looking for patterns, and identifying the stories that the objects told him. 

Today, as an architect in Centric’s Data and Analytics practice, Robinson does exactly that — but with data. And with the arrival of AI, he can now tell increasingly complex — and valuable — stories. 

“Until recently, I was skeptical about AI’s ability to help me do my job,” Robinson admits, even though he had written a paper in college about neural networks and artificial intelligence pathways.  

“But I began experimenting with Anthropic’s AI tool, Claude, and quickly saw how I could gain insights from data in hours that would have taken me weeks to do before,” he said. 

Robinson now realizes that his college paper, which he called “The Silicon Brain,” had envisioned AI’s future. However, in the mid-1980s, that future was not yet clear to him. 

 “I hadn’t put together that I had described where we are now,” he said. And for Robinson, where we are now is an exciting place to be — and a long way from where his career began. 

Uncovering Buried Data Secrets

Born in Germany, Robinson attended Bowling Green State University, where he met his “college sweetheart” and now wife of 35 years. After college, one of his first employers — a high-volume laser printing company — presented him with a technical challenge: Create annual employee benefit statements for IBM that include stacked, 3D bar charts in just three days. 

“At the time, Xerox’s existing technology couldn’t do that,” Robinson recalls. “So, in three days, I taught myself to write custom PostScript code, a predecessor to HTML, which allowed me to build the charts. I love solving problems and puzzling things like that out. We won the account.” 

One day in 1994, Robinson was on a chartered jet with the C-suite of a large grocery conglomerate. He was seated across from the CEO, who asked him what his role was at his company.  

“I told him, ‘I get you the information you need to make better decisions, faster,’” he recalls. The CEO shoved a pile of hand-drawn charts and graphs at him and said, “Can you get me this?” 

Robinson asked, “Would Monday be soon enough?” It was already Thursday, and the CEO told him the IT department had been telling him for two years that they couldn’t do it. How could he?  

Fortunately, Robinson had been collecting data from all the stores for a loyalty shopping program that was launching soon. He used that data as the base. 

By Monday, he had converted the hand-drawn charts to digital assets using Cognos PowerPlay, a multi-dimensional cube reporting tool. 

“That led me into business intelligence, and I never looked back,” Robinson said. 

Over the next three decades, he performed business intelligence for oil pipelines, retailers, manufacturers, healthcare and more. Each industry offered new ground for discovering and interpreting data artifacts. 

One client was a lawn tractor manufacturer. Robinson used the company’s data to build a sales dashboard for a skeptical division president. Sales were flat year over year and he was unimpressed, saying, “I already know that.” 

“I started drilling down into the company’s product lines with him and revealed that some retailers had stopped buying ride-on tractors in favor of self-propelled walk-behind mowers,” he recalls.  

“I then showed him that product margins for self-propelled mowers are 10-15 percent lower than riding tractors. So yes, the sales figures were flat, but more importantly, their profit numbers were lower because of the shift in consumer preferences.” 

In Robinson’s words, “When he saw that, his eyes about fell out of his head.”  

The division president immediately called the sales rep, got him on a plane to the customer HQ, and moved them to a higher margin mower series, recovering some of the lost margins. 

The buried story Robinson had revealed would have remained hidden without his careful data excavation and analysis. “Originally a doubter, the division president became one of our biggest supporters,” he noted.  

Finding a Home — And Data Insights — At Centric

Robinson capped off his pre-Centric career with a five-year stint at a major insurer, but by that time, corporate life had taken its toll. He was ready for a change. 

One day, someone at Centric reached out after seeing his LinkedIn profile. Centric’s Cincinnati team needed to find a new data and analytics local solution offering lead and Robinson was a fit. 

“I had received calls from companies like Apple and Microsoft, I’d never talked to a place like Centric,” he said. “I accepted the job, and at my first business unit meeting, I found everyone so welcoming, inclusive and committed to Centric’s values that I went home and told my wife, ‘I found my people!’” 

That was more than 12 years ago. As he had throughout his career, Robinson remained committed to his passion for versatility.  

“I have dedicated myself to not being pigeonholed,” he explained. “I’m uniquely qualified to play a lot of roles, and at Centric, I have the flexibility to do whatever our client needs me to do.” 

For one client, he returned to his grocery roots for another adventure in data archaeology. While analyzing customer pickup data and employee fulfillment patterns, his team discovered multiple issues in the data that skewed the KPIs for fill-rates and pick speed. 

“The employees picking the online orders were incented based on speed and developed creative workarounds allowing them to post very fast times,” he explained. 

“For example, they would collect items in their arms while walking the aisles, then return to the end of the aisle to scan everything rapidly, creating the illusion of lightning-fast picking. Those and other behaviors improved their performance scores but introduced customer satisfaction issues due to out-of-stocks and substitutions.” 

Armed with those insights, the team cleaned the data and created a highly successful mobile application that allows store leadership to manage the picking process in near-real-time. This increased fill-rates, reduced out-of-stock items, and identified areas of improvement for the store staff.  

Continuing to Evolve

Currently, Robinson is working on multiple business proposals and collaborating with advisory teams and various Centric centers of excellence on data analysis, modeling, and solution design. One significant focus involves using Microsoft Dynamics 365 to create a reporting solution similar to Centric’s Insurance Analytics Platform. 

The project further illustrates his commitment to versatility and growth. A year ago, he was focused primarily on using Databricks for data processing and using the Kusto Query Language (KQL) through Azure Data Explorer. Eventually, Azure Data Explore became Microsoft Fabric’s Event House. 

“I completed some Fabric Architect training last summer, and I started doing a lot of data mapping and Fabric Lakehouse work,” Robinson said.  

Working on a project with an international employee background check company, he was tasked to replace a proprietary third-party reporting solution on Microsoft D365 Finance and Operations (F&O). He had never used F&O, but he dove in and learn it on the fly with the support of Suzanne Love, from the MBA practice.  

“It was a tough three months, reverse engineering the reporting model into F&O without any documentation,” he noted. “Along with Shubham Kalkhuriya and Dhruwa Mishra from Centric’s India team, we successfully mapped more than 100 tables supporting 15 Financial reports.” 

And now, he has AI to help him. 

“I’m realizing that the key to incorporating AI into my Centric work is prompt engineering,” said Robinson. “If you can devise a prompt that asks the right question with the right amount of detail and examples, it can produce fairly accurate information.”  

For a new client proposal, the HR department needs to load sensitive employee data into Microsoft Fabric to replace several reports. Robinson was curious if AI could produce code to encrypt and decrypt XML streams directly into a Fabric Warehouse. He turned to Claude for assistance. 

“I devised a prompt that resulted in some very solid Python code that would have taken me days to write myself,” he said, cautioning that even with such results, you still need the human in the loop to validate the accuracy of the results. 

“It did take me a couple of rounds to craft the prompt to produce what I was looking for, and even then, I needed to review and validate the output. But it will continue to get better,” said Robinson.  

Still at Home with Archaeology

Archaeology still fascinates Robinson. One of his favorite shows to stream is a British archaeology program from the 1990s called Time Team. Currently, they are revisiting the famous Sutton Hoo site in England. He also enjoys cooking, “my love language,” he says, and has two children, a 15-year-old cat, and two granddaughters. 

“I love learning new languages and exploring cultures, especially Spanish, Chinese and Japanese,” he said, adding that he once traveled for 21 days in China with a friend from college.  

But whether he’s traveling the world or excavating data, he is committed to uncovering hidden truths that help organizations gain insights and make decisions based on data, not gut feelings, speculation, or assumptions. “Without data, it’s just your opinion.” 

“Data is everywhere now, but truly valuable insights are often hidden,” Robinson concluded. “I think my focus on staying open to learning new things and new technologies has allowed me to be the right partner for Centric’s clients, because it helps me see patterns others miss and ask questions others aren’t asking. That helps me find powerful stories that drive better business decisions.”