Adapting Voice Assistance for Business: The Crawl

The Vision for Intelligent Voice Assistance

Consumers have been using voice assistants such as Amazon Alexa, Google Home and Apple Siri for at least four years. I remember the days when it was extremely frustrating (and hilarious) to use Siri for anything useful. Recently, natural language processing and automated speech recognition have advanced significantly. In return, voice (or intelligent) assistants have gained huge acceptance in consumer applications – from home automation to ordering pizza. While this trend seems to have caught fire on the consumer side, it has yet to have any application in the business world.

As Smartbridge looked deeper into voice assistants’ expanding technology, we were convinced that there is an incredible opportunity to improve business processes with voice assistance. Smartbridge’s innovators are now leading the way to developing voice assisted solutions that enhance business efficiency from the executive suite to the warehouse floor. We will be sharing a series of blog posts highlighting our experiences and progression through our journey with voice assisted services for business.

Selecting the Platform for Intelligent Voice Assistance

Selecting a platform for our voice assistant was the first step. Our criteria were pretty simple: ease of programming and an open architecture. After considering all the leading platforms – including Google’s Assistant, Apple’s Siri and Microsoft’s Cortana – we selected Amazon’s Alexa.

Alexa does have some limitations; namely, it’s not available on traditional mobile devices, but rather in the Echo and the Dot. But, the fact that it is relatively easy to create skills and custom programs using Amazon Web Services’ Lambda, and lends itself to open architecture made it a winner for us. An additional bonus is that Amazon Alexa has lots of resources available to help people who are starting out on a similar journey.

Organizing our Thoughts

Smartbridge is an Oracle JD Edwards (JDE) company. We have experts that know the intricacies of its system and numerous clients that use JD Edwards as their backbone enterprise resources planning (ERP) system. Creating use cases with JDE was a natural fit. We gathered some of our leading JDE experts and identified various business scenarios that would benefit from being voice enabled and result in higher efficiency to the organization.

Conceptually, the business process would be enabled by the use of speaking commands to Alexa; Alexa would parse these commands and pass parameters via Lambda to a JD Edwards business service; and the returned information from the JDE business service would be translated back to the user by Lambda and Amazon Voice Services.

The First Crawl

We realized we needed a team with diverse skills to create the required data sets: Alexa skills, Javascript and Java programming, and knowledge of JD Edwards business services and system. With a small team of part-time members, we set our sights on our first objective: to prove that we could make a complete “round trip” from Alexa to JD Edwards and back again with verbal response.

We selected a relatively easy JD Edwards business service that would output the name of a contact from the address book. This business service did not meet the specific requirements of business scenarios we wanted to build out, but it did allow us to confirm the technology’s capability. Since the AWS Lambda function was outside our lab environment, we had to ensure that the JD Edwards business service endpoint was made available outside the domain. After a few minor hiccups, we finally achieved end to end integration. The test went as follows:

User: Alexa, open Address Book Manager.

Alexa: Address Book Manager. What address book ID would you like to look up?

User: Look up address book id: 20.

Alexa: Address book id: 20 belongs to Universal Supplies. What other address book ID would you like to look up?

User: Exit.

Next Steps

We finally learned how to crawl! Armed with the skills we gained with the successful end to end Alexa-JD Edwards point of contact (POC), we were ready to build Alexa’s skills for a few real-life business process scenarios.