Listen on iTunes  |   Listen on Spotify  | Listen on Google 

Episode  Summary:

In today’s episode, we will speak with Manny Bernabe and discuss the current landscape of AI, how to get started implementing AI solutions and what organizations should be doing today to set them up AI success in the future. Manny is the founder of and has 10+ years of experience creating and deploying AI & Machine Learning solutions and products to industries from financial services to semiconductor manufacturing. He helps innovation leaders build AI & analytics products/services while avoiding R&D failures.

Top 3 Value Bombs:

  1. If your organization want’s to be in business in 10 years, then you need to start prototyping/creating AI solutions
  2. AI should be top of mind when designing modern data architectures.
  3. Start collecting data today for the questions you want to be able to answer three years from now.

You’ll Learn:

The Current Landscape of AI / ML is Immature

Many organizations are struggling to get basic models off the ground. Many tools are making it easier to and lowering the skills required to perform AI/ML than 4 years ago. Overall the industry is lacking standard processes to explain how the AI is working in particular with deep learning / neural nets which is very important for industries with tight regulation like financial.

Two common misconceptions organizations have when it comes to starting new AI projects?

  1. Not getting started – If you want to be in business 10 years from now and not get disrupted then you should start experimenting and prototyping with  ML and AI today. Start to learn how it works for your industry and the use cases that will work well. Start building the knowledge base within the organization and know the types of historical data you will want for questions you want to answer in three years. 
  2. Getting started in the wrong way.  – Don’t build out a large AI/ML team from the start, start small and learn the specific roles you need on the team. There are many specific specialties within the ML space, for example, a computer vision expert vs a natural language processing expert. Sometimes you do not know what roles you will need until you get further along in the prototyping process. 

Questions a data leader should be asking when spinning up a project to implement an AI solution:

  • What are the main value driver for the business today? Can we streamline that? 
  • What’s going to be driving value two years from now?
  • Where is our expertise?
  • What data do we have? 

AutoML is not mature enough for non-data scientists to be successful yet…  But AutoML tools are trending in that direction. If your a data scientist it can greatly speed up your prototyping discovery phase. 

What’s been your biggest lesson learned when building out an AI solution?  You must have a strong data infrastructure in place and that is storing the historical data you may want in the future. Many organizations are not thinking about keeping data for potential AI solutions if they do not current AI solutions but they should be.  

What’s next for the AI industry over the next 2-5 years? 

  1. Companies will continue to struggle to do AI the right way
  2. Technology giants(FAANG) will continue to disrupt the space
  3. Companies not able to execute on AI solutions will most likely fail in 10 years