Top 5 Burning Questions on Business Applications of AI

 | 

We all have burning questions around AI and the new business opportunities it presents. At Solvvy, we are fortunate to have Tom Mitchell as an advisor and our in-house expert. Tom is widely recognized as a pioneer of AI & machine learning and wrote one of the first books on machine learning (aptly titled “Machine Learning”). He previously served as Chair of the Machine Learning Department at Carnegie-Mellon University, where he continues to teach today.

We recently sat down with Tom and had him answer our top questions.

Q: What are the benefits of introducing AI-powered solutions to my business today?

A: AI is a huge transformer of business. I believe that companies that adopt AI earlier will have a strong competitive advantage. Today, companies have no shortage of data and AI is able to process this data at scale to unlock actionable insights. This can be identifying areas of automation or trends across your customer base that you can use to take strategic action. The sooner you get these insights, the sooner you can implement initiatives that will set you apart.

Q: When it comes to AI, how should I think about when to build something in-house?

A: Build when:

  1. It’s proprietary to maintain competitive advantage: If you’re building something that differentiates your core product, build it in-house. A robust and effective AI-powered solution is not something you can build overnight. Be ruthless on deciding what to build–if a project isn’t central to your business mission, it likely isn’t the best use of your team’s time.
  2. You have sufficient team resources: You will need dedicated resources not just to build, but to maintain, train, and improve your initiative. Machine learning solutions require sizeable engineering infrastructure–make sure you have a team in place for this ongoing commitment.
  3. No vendors can meet your requirement: Build if you’ve done your research on vendors and came up empty-handed. Don’t build it yourself unless you have to. Using an outside vendor means you can leverage their expertise and existing infrastructure, thus saving you time, effort, and money. When evaluating vendors, push for customer references and case studies to make sure they have successfully deployed their solution to companies before.

Q: It seems like every company claims they use AI & ML in their product/service. How do you evaluate how “good” a company’s AI is?

A: How “good” a company’s AI is comes down to human talent, data, algorithms, and computing power. Given that computing power and algorithms are mostly commodities, data and human talent are the big factors. There are two key differentiators: 1) the relevancy and quantity of the data used to train the system and 2) how your human talent built your system. With these two in mind, it’s about a company’s ability to accumulate the best data and the best AI talent to solve the problem at hand. Ask companies how many customers they have (quantity of data), how similar the customers are to your business (relevancy of data), and how large the machine learning team is (how the system is built).

Q: Which companies do you think are leveraging AI well?

A: The obvious examples are companies such as Amazon, Google, and Apple who are investing heavily in AI. I’m also excited about innovative and younger companies that are tackling spaces ripe for AI-powered automation. I’ve been impressed with Clara Labs and x.ai who both have an early lead in leveraging AI for scheduling. Being early to market means they have a faster accumulation of relevant training data which makes it harder for new competitors to catch up.

Another company I’m excited about is Solvvy. I joined as an advisor because the customer support industry has a large opportunity for automation to improve how customers are serviced. Solvvy has an impressive customer base across industries, which means their training data is comprehensive and can provide better customer service across verticals. In addition to offering self-service, Solvvy is expanding to deliver business insights and automated workflows that will continue to drive significant value for businesses.

Q: What do you think makes Solvvy unique?

A: What makes Solvvy unique is its approach to the challenging problem of understanding user intent from both a technology and design perspective. The core technology they are building and their approach to training data are best-in-class and the results speak for themselves–I haven’t heard of an average 25% self-service rate anywhere else. From a design perspective, Solvvy provides a single interface that is intuitive and simple for all contact channels and that no other companies are doing that well today.

Have more burning questions on your mind? Send us a note at maggie@solvvy.com and we’ll include it in our next burning questions post!