The Dallas Data Science Conference for 2017 focused on the three key areas of technology, innovation, and career for a large audience of students, industry professionals, and entrepreneurs. The event was hosted by the Data Science Association and took place on the massive UT Dallas campus. After navigating the complex parking situation, attendees gathered in the auditorium to hear from a prestigious lineup of speakers from enterprise analytics pros to academicians.
NVIDIA, Data Application Lab, and TechTinx were just a few of the sponsors supporting the event. These organizations were eager to present the next generation of data scientists with a glimpse of the pathway leading into the business world. For some, this could include taking one or more of the many courses offered by the Lab, from Big Data engineering to Natural Language Processing. UT Dallas also stepped up to the plate by providing the venue for presentations and plenty of space in the Science Learning Center for students and mentors to network before and after the conference.
Agenda highlights the bridges that must be built
The panel discussion at the mid-point of the day brought together a mix of professionals from various data science disciplines to talk about bridging the gap between academia and industry. According to panelist Khusro Khalid from Research Now, “The audience was made up primarily of students looking for advice.” With things changing so rapidly in the field of data science, it was not surprising that those currently in training would question where the knowledge they gained in the classroom might fit in the real world of business. With companies giving preference to people with actual experience in data science projects in a business setting, students were understandably concerned about how to break into the job market without prior experience.
Andrew Savage, Data Science Career Advisor at Metis, recommended that students engage in projects during their coursework that demonstrate their ability to get results. “Hiring managers want to meet people who are creative problem solvers, people who are determined. If you can explain how you solved a problem, why you chose the methods you did, and why it’s relevant to a challenge in their business, that’s what they are looking for.” Students were encouraged to choose educational opportunities that would allow them to demonstrate capabilities rather than just presenting a certificate or diploma to potential employers.
The new polyglot data scientist must speak the language of business
Juan Vasquez, Communications & Data Analyst for the Operations Innovation team in the LA Mayor’s office, called on technologists to learn the language of business stakeholders. Getting things done with data science requires the ability to craft a story out of the numbers and statistics. “When you talk to a stakeholder, you have maybe thirty minutes in a meeting and only ten minutes to talk about the technology. You have to attach tech to a very clearly verbalized business problem.”
Sai Devulapalli, Data Analytics Product and Business Leader at EMC/Dell, spoke to a similar issue. His talk explored the learning curve business decision makers go through in embracing analytics. In Sai’s experience, stakeholders want to know the ROI, time to completion, hidden costs, and how the technology could change business processes. Even more important with advanced analytics is the issue of trusting technology to shape decisions. Organizations tend to start with small, predictive analytics projects that focus on reducing risk and cost. When ROI is captured from conservative projects, companies eventually become comfortable launching more mature initiatives for revenue generation and using prescriptive analytics that actually drive business decisions across the organization.
Speakers speak out on favorite topics
While the student population in the audience had much to learn, the professionals giving talks at the conference also appreciated the opportunity to glean insights from their peers. Dr. Meltem Ballan’s talk caught the attention of attendees for featuring the most unusual data source—bread mold. But her detailed explanation of image processing was the main takeaway for speaker Brad Taylor from Omnitracs. It spoke directly to a need within his organization. “Image processing is a problem that comes up increasingly in transportation applications.” He pointed out that being able to interpret what’s happening with street signs and other vehicles in real-time based on a video feed is a significant puzzle, but one the transportation industry would like to solve.
Data Skeptic Kyle Polich’s talk on the API economy drew praise from several of the other presenters. Data Scientist Neeraj Madan said the concept made a lot of sense when considering how data science will evolve in the future. “It’s about APIs making life simpler and reducing redundant or repetitive work.” Kyle himself found it interesting to see what was happening in related fields that will impact his own area of expertise, “It was good to catch up on NVIDIA and what they can do in the deep learning space. There’s big interest in GPU technology right now.”
Expert speakers offered something for everyone
Some of the presentations were readily accessible to those with little background in data science. But others took the topic to a much more advanced level. Brad mentioned, “The final presenter, Vibhav Gogate from UT Dallas, made some very interesting points around comparison of training sets. It was definitely at a good level for masters and PhD students. I was a bit blown away.”
For non-technologists in the crowd, it was an opportunity to be introduced to many new ideas at once. Gabe Bautista, a digital marketing consultant, spoke his mind. “I like to get a feel for things that aren’t my primary area of interest to gain context. I was aware of some of the technology, but not the volume or the speed at which it’s operating now. It would take days just to go through all the different platforms that were mentioned.” Indeed, the one-day conference gave one takeaway for all attendees—this is just the beginning for Big Data Science.