Chatbots and AI Virtual Agents

It’s time to treat Artificial Intelligence like what it is: here to augment human productivity, not replace it. The Sparkcentral Virtual Agent Framework considers chatbots “digital employees”, collaborating with your live agents.

Get a Demo

AI should not replace humans, but complement them

Leading analysts estimate that by 2022, 70% of customer interactions will involve technology such as chatbots, with 20% of all being completely handled by AI.

However, they also estimate that 40% of interactions will still involve humans.

Customer service is sometimes about advising, consulting and providing guidance to customers. It can require a level of emotion, or empathy that can only be achieved by a human.

At Sparkcentral, we therefore believe that AI will have the role of working side-by-side with humans to augment, not replace them.

Three key capabilities any chatbot needs

Virtual agents that help with serving your customers need the same capabilities as live agents to reach full productivity:

Lead a conversation

The core ability is of course to engage with customers on the channel of their choice. Any chatbots deployed on Sparkcentral can be deployed to have full conversations with customers on any channel you are managing through our platform. Adding a Virtual Agent to a new channel takes a click of a button.

Manage a conversation

Virtual agents also need to be able to assign a conversation to other virtual agents, or back to live agents to take over. Furthermore, they need to be able to set key conversational attributes such as topics or tags, as well as leave notes for other agents. Sparkcentral enables all of the above, and more.

Conduct transactions

Finally, virtual agents need to be able to book rooms, look up balances, change a reservation, and do all the other things in your backend system that your live agents are doing. Virtual agents deployed on Sparkcentral retain all of these capabilities, for full productivity gain.

Deploy your own Tier 1 Chatbots with Sparkcentral

With Sparkcentral, you can easily configure simple chatbots for tasks such as asking pre-qualification questions or providing out-of-hours — similar to the job your tier 1 agents are doing today in your customer service organization.

Configuring your tier 1 chatbots is as easy as providing the questions you want them to ask, or the messages you want them to provide. Choose:

  • Whether they should come in at inception point to ask prequalifying questions on behalf of your live agents
  • Whether live agents should be able to delegate a conversation to a tier 1 chatbot mid-dialogue
  • The business hours for auto-responding when out-of-hours
  • The channels you want to deploy them on

BYOB (“Bring Your Own Bot”) for full AI and NLU-enabled chatbots

Sparkcentral lets you connect any AI-powered chatbot you create in any platform you choose to any messaging channel you want.

Sparkcentral recognizes that many organizations already have introduced virtual agents to their digital customer service operations. Rather than forcing you to re-build your chatbots on another proprietary platform, Sparkcentral integrates with any chatbot vendor through a versatile API.

Don’t have an AI partner yet? Talk to us, and get introduced to one of several chatbot partners specializing in different verticals and different regions of the world.

Message Us

When to bring a virtual agent into a customer conversation

Sparkcentral lets you employ as many virtual agents as you need, for different uses cases and “skills”. The approach allows human handover at any point in the dialog, even during a conversation – so your customer always gets the best possible experience.


The Virtual Agent is assigned a new conversation. They can resolve the conversation or hand it over to a live agent at anytime.

Use cases:

  • Identify and authenticate the customer
  • Determine the reason for contact
  • Route to a right-skilled agent
  • Answer FAQs


During a conversation, a live agent can pass the interaction to a Virtual Agent to perform simple data collection and basic transactions, then claim the dialog back.

Use cases:

  • Collect data for appointment bookings or reservations
  • Collect information to complete form fields used in business processes
  • Gather credit card details


Once a live agent finishes a conversation, they can pass the interaction to a Virtual Agent for follow-up actions.

Use cases:

  • Conduct CSAT surveys
  • Ask for a follow-up date to continue a business process
  • Send a digital voucher or coupon

A sample Sparkcentral dialog employing live and virtual agents as a team

Business value propositions

What will adding Virtual Agents do for your digital customer service operations?

Reduce cost

Support your live agents with automation that takes away the easy and repetitive side of customer service

Increase productivity

Free up live agents for matters that require human empathy, judgement, intelligence

Extend business hours

Use across all Messaging channels and extend your business hours to 24x7x365

Improve customer experience

Answer simple questions faster and with less customer effort — and no wait times

How we engage with you to deploy your first virtual agent

Phase 1: Working towards “Day 1” of your new digital employee


  • Collect top 30 contact reasons (“intents”)
  • Identify which of these are automatable with a standardized
    dialog and workflow
  • Select list of top 1-3 to start with
  • Make design decisions (persona, communication style (menu-based,
    natural language), entry points (Inception? Delegation? Conclusion?), …)
  • Select implementation partner


  • Create dialog design
  • Implement dialog design, incl. backend integrations
  • “Hire” Virtual Agent (= add to Sparkcentral system and provide technical
    integration; start including VA in workflow)
  • “Train” Virtual Agent on the job (=monitor performance, improve responses)

Phase 2: Expanding the capabilities of your digital employee


  • Select next 5-10 intents
  • Design Dialog Flow and Implement
  • Go from menu-based to full NLU
  • Build out specialized Virtual Agents for smaller tasks
  • Start expanding on delegation scenarios, having live agents increasingly leverage virtual agents during conversations

Learn More

Chatbots in the contact center won’t replace live agents — our product manager gives two reasons:

1. Chatbot applications are only as good as the data they leverage
2. Importance of human interactions

Chatbots contact center live agents

Request a demo

Engage in a digital dialogue with your customers and resolve 100% of their questions and issues fast, securely, without ever overwhelming your agents. When you put Sparkcentral to use you receive more than a proven platform, you also get expert guidance to make your team the most productive it can get!

Message Us