AI Trends interviewed Barry Coleman, CTO, Agent.ai

Q. Tell us about Agent.ai?

A. Agent.ai is a customer service platform designed to help businesses connect in real-time with their customers wherever they are, whenever they’d like and in the communication channels they prefer. Agent.ai provides a robust standalone CRM or integrates with customers’ existing platforms like Zendesk, Desk, Intercom, and Freshdesk, to offer their support agents a powerful and consolidated dashboard view of customer event data, analytics, session information, transactions and more, in order to support better case management and streamlined customer issue resolutions.

With the integration of Agent.ai, CRM software can leverage machine learning to understand agent and customer conversations and soon begin to both anticipate and suggest accurate responses for the majority of customers’ inquiries. This allows customer support teams to focus on issues that need more dedicated attention, while providing the majority of customers with immediate resolutions without the need for human interaction.

Q. Can you tell me about the core features of your Agent.ai product that make it different?

A. At Agent.ai we have a unique approach to implementing and integrating our AI platform, with the aim of increasing automation over three stages:

  • The first stage consists of the AI engine passively monitoring agent interactions and company logs, as it learns to suggest and mimic the best agent responses.
  • In the next stage, called Co-Pilot mode, customer service agents are presented with increasingly accurate machine-learned and agent-taught responses for their customer inquiries directly in the Admin Console, enabling a simple review-and-send response process. The AI in Co-Pilot mode is acting more like an assistant. This mode leverages a mix of AI technologies to analyze customer and agent interactions, ticket resolutions, company knowledge base, FAQs and custom-selected events to anticipate and surface a logical response for the agent to forward along to the customer. The goal with our approach is to establish a reasonable set of intent, make the system as “smart” as possible, and create maximum efficiencies for the agents first — before moving on to full automation.
  • The final stage is called Auto-Pilot mode.  Over time, Co-Pilot responses increase in accuracy, to a point that allows companies to let Agent.ai become the first responder to customers’ frequently asked questions. This allows companies to provide immediate responses without agent intervention. The software includes logic parameters to only respond to inquiries when it has extreme confidence in the appropriate response, and otherwise routes questions to human agents.

Q. How do customers get started with Agent.ai?

A. Agent.ai offers a simple setup and onboarding process, enabling most customers to deploy within an hour. Additionally, we offer unlimited syncing of customer support ticket data to properly train our AI engines. Furthermore, unlike other solutions, Agent.ai enables organizations to service their customers in seconds, mirroring the response users have come to expect from messaging apps, without sacrificing the high-touch experience that customers are accustomed to.

Q. Do you see AI being integral to customer service going forward?

A. As the technology improves, you can bet that a lot more companies will start integrating AIs into their teams. According to research agency Gartner, “By 2020, customers will manage 85% of their relationship with the enterprise “without interacting with a human.” Machine learning just makes sense as a CRM tool. Over time, such a system can learn best practices for a wide range of company-specific issues. The responsive AI assistant can get back to customers with an answer (or question) in a matter of seconds, as opposed to days. It can make existing CRM teams more effective, and offer options for companies who need more flexibility or can’t afford to hire a large team.

For more information or to get a demo, visit Agent.ai’s site, or contact demo@Agent.ai.