Big Data Profile: Catalin Ciobanu, Carlson Wagonlit Travel

Catalin Ciobanu

Tenure at current job: 3 years. Prior to entering the business world, Ciobanu spent 10 years doing research at Fermi National Accelerator Laboratory (Fermilab).

Career accomplishment of which I'm most proud: Leading the big data transformation within the company, where I demonstrate to multiple groups how big data and advanced data analysis can help them. As part of this effort, I try to enhance the way my colleagues approach problem solving within the company.

Decision I wish I could do over: [There were] a couple places I could have taken more risk. Prior bias is another one of my faults. For example, the concept of [the] real-time dashboard wasn't that big in my mind. But once we built it, a lot of people came in with ideas other than what I had in mind ... I was wrong to think I knew the use cases beforehand.

Most important career influencer: Within my career in the business world, it would be my current boss. He understood my value, and recruited me to the team, before I did. Also, my dad was a physics teacher who influenced my way of thinking in a thousand dinner conversations and debates.

One thing I learned during my scientific career was to ask the big questions. The most fundamental questions are the hardest to solve, yet everybody has an idea what the answer is. By putting a framework around it, you let data guide you to the right answers.

Current top initiatives: 1. To group all data sources in order to reconstruct the "trip experience," from dollars spent to the amount of stress a traveler experiences. There is also some data we do not have and we are hunting for. 2. To exploit real-time capabilities, maximizing what we can offer our travelers and our suppliers.

Most disruptive force in my industry: In my little business intelligence universe, the most disruptive force is the explosion of powerful analytical tools, many of them open source. For 20+ years, IT's goal was to rationalize and unify software. [But] now, with cheap hardware and open source software, we're moving back to customization to catalyze innovation.

One thing I'm looking to do better: Scale what I'm doing. To pick up more momentum it will take both more people and certain new processes.

Most common reason big data projects go wrong: Access to data. Big data projects have stringent requirements for data access: fast, repeated and full datasets. When I started, we lacked a framework for this but now we are on the right track.