Lessons from AI & Big Data Expo North America 2018
Last week, the Ople team exhibited at AI & Big Data Conference & Expo North America 2018. As I mentioned in our previous event post “Plug and Play Fall Summit 2018 Recap,” we have been listening to our customers, and the team has been working hard to bring AI transparency, including model explainability, to the Ople platform. We are strong advocates of AI transparency (Read Pedro’s blog posts on Transparency), and customers love the updates!
We have an insatiable appetite for feedback. We spoke with companies both small and large; leaders across the organizations - Data Scientists, data science team leads, CTOs, and CEOs; and companies across industries - including retail, manufacturing, CPG, and eCommerce. Across all of these conversations, we learned three big lessons.
1. Time to value matters most
Companies of all sizes shared the frustration with limited access to Data Scientists. Time and resources are limited, and when projects take months - fewer teams get the opportunity to explore new ideas. Surprisingly, this frustration seemed most acute in companies that have data engineering in place.
For example, a CPG executive shared that he is desperately looking for a way to accelerate his data science team. He was very frustrated because his team already has a number of data sets ready to go but cannot get access to a data scientist.
We learned how critical AI transparency is to internal buy-in at VB Summit (Pedro dives deeper in this article). Not unsurprisingly, we heard the same thing at AI Expo. AI is very powerful, but teams struggle to implement AI models because internal stakeholders do not understand or trust how AI is making the predictions.
A business analytics team lead shared with us how transparent AI is lowering the barrier to data science project adoption. His approach was to have a data engineering team and his analysts prepare the data and use Ople to build an AI model. Our AI transparency tools allowed the data science team to validate the model and explain the outcomes to other stakeholders.
As more companies invest in AI, customers are finding opportunities for AI across their organizations. Consulting companies, especially, were asking for across-the-organization-access to AI because they work with multiple teams, organizations, and industries solving very different problems.
Accessibility resonated strongly with companies that are operating in multiple verticals. For example, a data science team lead at a European consulting company was delighted to learn that Ople is industry-agnostic software, not a point solution. She shared that her team has been performing a rigorous screening process because their models would only work under certain industry specifications. By adopting a tool that can be applied across industries, she pointed out how her team can work with more clients and gain competitive advantage.
In fact at his keynote on “AI for Everyone,” Pedro spoke openly and honestly about how to overcome these issues based on his past experience as a Chief Data Scientist and now as CEO of Ople. After the keynote, people lined up to speak with him to thank him for very thoughtful and honest advice, and to ask questions about their experiences. This ad-hoc group discussion became a crowd of its own and we had to move to a separate room.
Based on the feedback and response from the market, it feels like we’ve hit on a key pain point! When we demonstrated the new transparent AI release, it was received with great enthusiasm on the floor. If you would like to learn more about Ople and transparent AI, contact us!