Sparkcentral virtual agent framework blends AI, live agents

Bring Your Own Bot, Sparkcentral's virtual agent framework for chatbots, blends live agent and virtual agent assistance to augment the customer experience.

Sparkcentral has rolled out an AI virtual agent framework that allows companies to combine live and virtual agents in contact centers. The Sparkcentral virtual agent framework, dubbed Bring Your Own Bot, complements the Sparkcentral Messaging Customer Service Platform, according to the vendor.

With Sparkcentral's Bring Your Own Bot framework, enterprises can retain their existing AI and natural language processing investments instead of rebuilding chatbot logic in a new tool; companies can connect any virtual agent on any platform via any messaging channel.

The virtual agent framework allows the deployment of chatbots on all chat, social and messaging channels supported by Sparkcentral's platform, such as Twitter, Facebook Messenger, WhatsApp and more.

With the Sparkcentral virtual agent framework, live agents have the flexibility to delegate simpler tasks to AI-powered agents mid-dialogue, according to the vendor. And virtual agents can pass the conversation back to the same live agent.

According to Sparkcentral, the ability to pass the conversation from live agent to chatbot back to a live agent -- what the vendor calls AI Team Play -- frees live agents up for more complex tasks while retaining agent availability across all communication channels.

Sparkcentral claims that virtual agents should not be a last resort to customer communication, but instead part of the process from start to end. There are three phases of the system's customer experience management, starting and ending with virtual agents: inception, delegation and conclusion.

Inception consists of a virtual agent gathering basic customer information, determining the reason of contact, answering FAQs and routing the customer to an appropriate agent.

After a live agent chats with the customer, the agent delegates simple tasks and transactions to a virtual agent -- such as collecting credit card information, booking appointments or filling out forms -- and then rejoins the conversation.

At the conclusion of the conversation, a live agent can pass off to a virtual agent once more to conduct a survey, schedule a follow-up appointment or other basic finishing transactions.

Earlier this month, Microsoft created a similar bot framework that supports multi-domain and multi-agent conversations; Microsoft plans to integrate this framework with its voice assistant Cortana.

According to Gartner, 25% of customer service operations will integrate virtual customer assistant or chatbot technology by 2020, compared to less than 2% in 2017.

To accommodate customers without an existing AI technology or preference, Sparkcentral will be partnering with several chatbot design agencies and conversational AI technology providers that are certified against the vendor's platform.

Dig Deeper on Customer service and contact center