Pittsburgh AI Works Launches INFOpgh Executive Seminar

Data fabric… knowledge catalogs… marketing science… digital transformation… These were some of the themes energizing the invite-only audience at INFOpgh, the inaugural AI seminar orchestrated by the DES data science division, Pittsburgh AI Works.

“The aim of the event was to bring regional business leaders into a high-value conversation around data and AI,” explained event organizer Scot MacTaggart, Chief Innovation Officer at DES. All too often, when these concepts are floated in the boardroom, they are not distilled into executive language around strategies and broader business objectives. This creates a disconnect between those mandated to perform and those expecting results.

The sequence of presentations at INFOpgh, aptly held at the Energy Innovation Center in the Hill District, led the listener through the current conversations around data science and AI — from the wonderous to the pragmatic.

Ryan Hoffman, director of CMU’s CREATE Lab, opened the seminar with dazzling visualizations of global phenomena — and challenged the audience to interpret the data in paradoxical ways. “Data gives you information, but you must thread it into a story,” he mused. The key is to choose your data sources carefully and ensure they justifiably contribute to the topic you are exploring.

Brian Dragunas of IBM, delved into data sources. He illustrated the methodology of collecting disparate data sources, applying processes to ensure data integrity, converging them to create a knowledge catalog — which then becomes the backbone for your data science and AI applications. Oftentimes, it’s the banal errors that corrupt data: the misspelled names, the duplicated information, the lack of protocol, the disarray of file locations. Much of this messiness can be solved through choosing the right tools for data creation and data integrity. Rather than the drudgery of manual data cleanup, advanced automation tools can right-fit your data in a fraction of the time.

While data integrity is imperative, “interpreting data is both an art and science” espoused Leah Jakaitis, owner of consultancy Totem Analytics. With towering achievements under her belt (Leah headed marketing science at Acrisure — yes, that same Acrisure who clinched the naming rights of Heinz Field), she shared several key principles behind marketing data, namely, how to calculate Customer Acquisition Costs (CAC) against their Lifetime Value (LTV).

Curt Wadsworth, patent attorney and developer of Dorothy AI, a competitive analysis tool for intellectual property documentation, showcased the possibilities of transforming unstructured documents into a critical knowledge bank. What if documentation could be scanned and intelligently tagged? What if you could use natural language to ask questions of your knowledge bank, and get the right answers? What if you could reduce research time from hours to seconds?!

Lastly, Sriniwas Silam, Director of AI Innovation at Pittsburgh AI Works, took to the dais and brought it home. Data can no longer be a strategic initiative — it must be the strategy itself. Any initiative that a company faces can be accelerated through the adoption of digital transformation. “It’s a mindset. It’s the new way of doing business,” Sriniwas affirmed. The most forceful inhibitors to data adoption are what he termed “Big Bang Syndrome” — wanting to take on too much, and “Zero Bang Syndrome” — approaching digital transformation haphazardly and ineffectively. Instead, define a business goal, get educated on the plentiful practical solutions in the AI toolbox, and get started.

The next executive session of INFOpgh is slated for April 2023. Subscribe to the mailing list for seminar update. To learn more about the practical AI offerings of Pittsburgh AI Works, please visit www.pghaiworks.com or email info@pghaiworks.com