Next Major Disruption In Computing; Voice and IoT

Blog:

First came the Babbage engine. Then simple, yet obscenely large, computers used for World War II. Chipsets less powerful than the phone in my pocket putting a man on the moon. Fast forwarding through all the technological advancements, we are now seeing an explosion of computing power. The technology isn’t relatively new, but Natural Language Processing is making headway into the homes of the consumer and businesses alike. Salesforce’s Trailhead on Alexa puts it plainly, “At Amazon, we believe voice represents the next major disruption in computing.” 

As of late, I’ve embraced Alexa. It’s connected to almost everything in my home. Not too long ago I stumbled upon some ideas of integrating Alexa with Salesforce. I saw there is surprisingly little when it comes to open source or popular ideas of integrating both systems. I followed and completed the well-written Trail on Alexa Development. Regrettably, I do not have the strongest background in Machine Learning, but I was pleasantly surprised how straightforward the development seemed to be. There are some learning curves if you are not familiar with REST, NodeJS, or AWS Lambda development. 

The question I kept asking myself as I performed each of the Trailhead tasks was, “Why should I be using this in Salesforce?” I originally thought of Alexa as an overly intelligent timer or music player. After completing the Trail, I saw it can help in simplifying call centers, accounting, warehouse management, Salesforce administration, and many more every day tasks. Consider the following questions you may be able to ask Alexa soon, should your business integrate with it: 

• “Alexa, move 6 of item X123 to location B”
  “Alright, I moved 6 of item X123 as an inventory transfer” 
• “Alexa, which of my Opportunities currently shows the highest profit margins?”
  “Out of your open Opportunities, it looks like your Opportunity with Cloud Computers has a profit margin of 32% or $40,000” 
• “Alex, what is the Accounts Receivable number of Universal Containers?”
  “Universal Containers currently owes $10,000” 
• “Alexa, deactivate John Smith”
  “John Smith deactivated” 

Alexa can accomplish the above and more. The Trail gave great insight into how you can leverage the power of AWS Lambda with nearly anything computer-based. I’ve come to view Alexa more as a personal/business assistant.  

When developing what Amazon calls Skills (or “Alexa App”), it is crucial to understand how Alexa converts the spoken language into your Salesforce business logic. Below is a general flow: 

1. Alexa does the heavy lifting by parsing the audio of the user into keywords. (see below with examples of what the Skill is expecting) 

2. The name of the Intent, above example is “UpdateOpportunityStageIntent”, is then one of the functions defined in your lambda (below – secureHandlers.js from the Salesforce Trail). It has seven utterances it is expecting for the same logic. The second image below is called in the salesforce.js file to update the chosen field. In this case, the Stage. 

The example provided by Salesforce in their Trail sets a strong foundation, so you don’t have to reinvent the wheel on structuring business logic. If you have an idea on utilizing Alexa with Salesforce, it can be nearly as simple as defining steps like you may normally do in Apex.  

Earlier I mentioned that I see some limitations. For one thing, adoption is a major obstacle. My parents tell me I didn’t learn much from Hal 9000 in the Space Odyssey. Another issue is dealing with lack of a screen entirely. My entire career up to this point revolved around a screen and now I must rethink software development. Amazon makes it clear, in order to make Alexa Skills successful they must feel natural and fluid. If the Skill requires a manual or significant training before use, it probably won’t succeed. Despite these issues, Voice-powered technologies, such as Alexa, provide the next level of computing opportunities that will no doubt become essential for everyday businesses.