How “Human in the Loop” AI Delivered Improved Customer Outcomes and Agent Productivity

More chats per day
0 Minute
Faster communication completion
Higher issue resolution rate

When Knoah, a leading business process outsourcer based in Nevada, Honduras, Guatemala, and India, wanted to empower their agents to handle more customer conversations and improve customer satisfaction, they knew a simple, self-service oriented chatbot wouldn’t work.

Knoah regularly works with a Fortune 100 electronics goods supplier, handling customer care and technical support cases that are far too complex for a chatbot — the risk of even a handful of interactions going poorly is simply too high, and costly.

So they turned to Augment — a conversational AI provider who focuses on improving digital customer experience through “human-in-the-loop” technology that pairs humans and AI, making both more effective.


Consumers often turn to digital messaging channels to buy from and interact with the brands they love. Unfortunately, these brands are discovering that the enterprise isn’t designed to effectively support high-volumes of customer conversations.

This is obvious from looking at commonly noted customer pain points:

  • Getting reliable information in a timely manner.
  • The ability to connect with a real person instead of an automated menu.
  • Having a quality support experience with minimal waiting.

A key issue is that mission-critical customer and product data often lives in hard-to-access CRM systems and out of date knowledge bases far outside of the engagement platform agents use to interface with customers.

Even if agents do have relatively easy access to the right data, the sheer volume of available product and customer information makes it nearly impossible to correctly apply any of it in real-time conversation, particularly when agents are handling multiple transactions concurrently.

As a result of these data access issues, the customer experience is often delayed, impersonal, and dissatisfying. The agent experience is equally frustrating — they end up copying and pasting many of the same “cookie cutter” answers over the course of a shift, and are unequipped to respond to complex product inquiries and issues when they arise.


Knoah needed a technology solution that:

  • Put the right information at the right time at the fingertips of their agents.
  • Allowed the agents the ultimate responsibility to use or not use the recommendations.
  • Improved on its own with minimal upkeep or maintenance.
  • Ultimately enabled agents to resolve more customer issues in less time.

Knoah already used the LivePerson’s LiveEngage platform, so this made it easy to leverage Augment’s ChatHelper feature, which seamlessly integrates with and sits inside of the LiveEngage platform on the agent’s desktop. Using past chat transcripts and workflows, Augment developed a dynamic knowledge base, with an emphasis on an advanced machine learning component that could deliver a self-correcting system with increasingly predictive value.

ChatHelper uses AI as an intelligent assistant — ensuring the live agent and knowledge base function as a single organism. Armed with ChatHelper, agents are be able to:

  • Engage in multiple chats simultaneously, and be confident about the information they give.
  • Access recommended responses and product information without ever having to leave their engagement console.
  • Access relevant subject matter articles they can share with customers.
  • Onboard new agents quickly.
We're thrilled to work with Augment as one of our preferred AI partners. Their 'human in the loop' approach offers a critical AI-capability for enterprise organizations who want to leverage messaging as a strategic channel for improved customer experience and revenue growth.
Paige Twillmann, AI Solution Lead at LivePerson


To test the value of ChatHelper, Knoah divided agents into one group that used the tool and one that did not. As compared to their counterparts, the ChatHelper group:

  • Handled 13% more chats per day.
  • Completed customer conversations almost one full minute faster (AHT).
  • Had a 2.3% higher issue resolution rate (FCR).  

There were also qualitative impacts that ChatHelper was able to deliver:

  • Shorter ramp up time in moving new agents from training to production.
  • Reduced stress levels for agents, due to the ease of providing answers plus greater customer satisfaction.
    • This also created a reduction in overall agent attrition.
  • Consistent and “fresher” answers to customer’s most common questions.  

Augment’s “human-in-the-loop” approach combines the power of AI to structure large data sets and continuously learn, with uniquely human capabilities like context recognition and empathy. This proved to be a powerful tool in helping Knoah’s agents become more productive and improve customer outcomes.


Interested in working with us?

Share this