Let’s face it, Artificial Intelligence is the latest breakthrough technology that has all the excitement right now.  You probably can’t make it through a full day without reading an article on it (like this one) or hearing someone talk about it.  AI has been around for a while, but we have seen some incredible breakthroughs the past few years. Let’s dive into some interesting solution considerations for an Ai strategy.

There is no question that AI will change the industry and how we work in the future, but AI is like any other software.  We need to look at it to validate how it can work best for us and the value it can provide.

• Sizing.  There is no “one size fits all” or even a “one size fits most” AI technology that will cover everything you will want and/or need.  The best strategy will consist of leveraging multiple AI technologies.
  
  
• End Use. You will need to understand the various models to understand its intent to ensure it will fit your business need.  The promise of AI is very enticing but misunderstanding it’s nature can result in missed expectations and/or failed implementations.
  
  
• Scalability. Systems can perform extremely well when you have one person using it, but what about 10 people, 100 people, 1000 people?  Scalability is also not just how fast it responds, but how fast can you deploy and train your teams to use it.  There isn’t a lot of value in having a fast solution that is too complicated to use.
  
  
• Simple to use, but not simplistic. There are many powerful models that can perform basic AI tasks (example: text to image).  However, the real game changers go far beyond just generating images from text and we’ll talk about these possibilities in future articles.
  
  
• Security. Product data is one of, if not the, most important aspect of your business and security must be considered throughout the entire stack.  Application isolation, auditing, security detection/automation, and company managed authentication are just a few critical elements.
  
  
• Predictable Pricing. Many solutions in this space are based upon credits which require you to properly forecast your usage.  Until we can create an accurate AI to forecast your AI usage, credit-based licensing can and will quickly add up ultimately exceeding your budget.
  
  
• Legal Considerations. We need to ensure that users cannot easily/accidentally leverage a model that introduces legal risk to the company.  We also need to provide tools / auditing for legal teams so they can help ensure what AI models are being used and how they’re being used.