Neuraflash at Salesforce Trailhead DX 2017: In-Market AI Applications
Salesforce Trailhead DX was held in San Francisco June 28-29 this year to celebrate and inspire the many developers and partners that build on the Salesforce platform.
Dennis Thomas, CTO of NeuraFlash, participated in a panel of Salesforce Incubatorparticipants, demonstrating in-market applications for Artificial Intelligence (AI). The panel titled, AI for the Enterprise: Use Cases, was led by Mike Kreaden, Managing Director, THE INCUBATOR, and included Greg Lewis, CEO, Hilo Studios, and Jon Jessup, CEO, Cloud Conversion.
SALESFORCE’S ALEX DAYSON CALLS TODAY “THE AGE OF THE CUSTOMER”
In his Trailhead DX opening presentation Alex Dayon, Chief Product Officer at Salesforce, called today “The Age of the Customer.”
This is at the heart of what NeuraFlash does – using AI built on the Salesforce Einstein platform to dramatically and continuously improve the customer experience and increase engagement to drive revenue.
NEURAFLASH CTO, DENNIS THOMAS, GAVE AN OVERVIEW OF AI IMPLEMENTATION WITHIN ENTERPRISE COMPANIES
Dennis told the audience that NeuraFlash is a Salesforce consulting partner focused on the intersection of Salesforce consulting and AI implementation within Enterprise companies. The leadership team brings together over a decade of experience in machine learning and speech recognition to apply best practices in AI across the Salesforce product suite.
He said that as one of 13 organizations in the second Salesforce Incubator cohort, NeuraFlash uses the Salesforce Einstein platform not only to process data but to transform the data that is Salesforce “gold” to drive better value in two main areas:
- Customer facing: Think of chatbots, conversational interfaces that drive automation both to help get questions answered for consumers and to determine the right moment to engage the customer. From an interaction standpoint users can communicate in Facebook messenger, the website, or any other channel: the value here is capturing and learning from customer data to help to answer routine questions, direct users to relevant sales or service agents and complete tasks on behalf of the user.
- Sales/Service Employee [or Agent] side: Neuraflash provides appropriate recommendations relevant to the customer or employee on the digital channel – ensuring that any recommendation isn’t just a blanket offer that’s applicable to the whole population but targeted to that particular user, based on who that user is or previous purchases in their profile.
The Einstein platform is not just about being able to process a bunch of data, in this case, it’s about the user experience tied to the sales/service employee [agent]. The data doesn’t produce a number or set of labels, but something that’s intuitive to the agent so that they can provide a recommendation in the moment or provide the relevant experience to the end customer.
He finished up by explaining that implementation is a combination of the core technology that Einstein provides and helping the Enterprise both adapt that technology to their customer experience and understand how it impacts their business overall.
DENNIS ANSWERED QUESTIONS FROM THE AUDIENCE ABOUT AI APPLICATIONS
Q: Are we seeing A.I. used as an evolutionary tool developing for additional functionality or is it really unlocking some revolutionary change to solving certain business problems that we just couldn’t do before?
DT (Dennis Thomas): Looking at the data right now, machine learning is addressing operationalizing the ability to get to the better recommendation, score, better opportunity from the revenue standpoint.
The bigger take away is that judgement is still in the hands of the employee/user that has to make the appropriate decision. It goes back to human nature. The system gives you a recommendation but the user can choose whether or not to act on that recommendation.
So, the question is, can these systems evolve from providing better number crunching ability to driving better self-service capability? For one customer we work with, 80% of live chats are common questions that can be addressed. They aren’t worried about productivity from a contact center standpoint they just want their human agents to focus on the harder problems. Answering the same basic questions over and over again isn’t valuable to the employee or the business. Right now, our natural language processing chatbots are answering these routine questions.
The evolution of the technology, through the work of the Salesforce research team in areas like question answering capabilities, is pushing the envelope from evolutionary to revolutionary to drive some of that automation even further.
Q: How do you gather information on a new customer to build your database? Do you use something like Facebook that already has data?
DT: You can try to use information from existing channels like Facebook but you often run into various permissions and security. What we recommend is that you use conversations across your different channels whether it’s a phone call, live chat or email – what are consumers asking about? Those are sources of conversation that help define the “voice” of the customer in terms of stated problems or questions that they have. Take those sources of data to bootstrap your natural language model if you’re going for a conversational interface. Use the data you have if you have it. If you don’t, fake some data out to use as a baseline then refine the model as you collect real data.
Q: If you could dream up the roadmap across all of Einstein what would your next killer feature be?
DT: Next generation question answering capability. When you access a search engine and type in a keyword now, you get a list of documents and the burden is on the user to click the document to find the reference within a search result. The value of the Enterprise is that it has a lot of knowledge-based articles, data and human capital that is in some kind of electronic format. Can the Einstein platform curate the Knowledge articles not only to get you the right article but to answer the question in the voice that you asked it? For example, “how many bags can I bring on the plane?” I don’t want the read the article, I want the system to say “Two bags, Dennis, because you are a platinum member.” That’s all about question-answering as a capability and where AI is heading.
###
NeuraFlash is an innovative Salesforce Consulting Partner, headquartered in Boston with offices in the San Francisco and NYC area, deploying intelligent solutions powered by Artificial Intelligence, including conversational Einstein bots across the digital channels (Web Chat, Facebook Messenger, SMS, mobile apps), Salesforce Service Cloud and Salesforce Einstein® solutions, to provide compelling customer experiences and improve employee productivity. Our team includes Certified Salesforce Consultants, as well as A.I., Digital Channels and Contact Center experts with decades of industry experience. NeuraFlash is 100% committed to exceptional customer service. For more information, call 1-844-638-7235 or visit neuraflash.com.
