Conversational agent back office

This guide introduces the Conversational agent back office and its elements.

What is Conversational agent back office

The Conversational agent back office interface consists of two key components—transcripts and agents.

Transcripts

Transcripts are detailed records of conversations between customers and Conversational agent. Reading through transcripts helps you analyze the customer interactions with Conversational agent in detail.

Agents

An agent represents a Conversational agent personality that targets a specific business goal. You can create and tailor multiple agents with a unique goal, catalog, language, currency, and content settings. Agents show distinct behaviors aligned with their purpose. The following are a few examples:

• Fashion sales agent: The fashion agent speaks in a trendy, youthful tone for the fashion-savvy audience. It sells trendy clothes or accessories using your fashion catalog.
• Czech skincare agent: The skin care agent for the Czech Republic region speaks Czech. By acting as a skin health guide, the agent builds brand loyalty among your customers.

Launch use cases with agents

While an agent targets a single goal, it adapts to different use cases on your website. The agent's core purpose remains the same. You can tailor its implementation to remove the friction points for each use case. For instance, a sales-focused agent might assist users on product listing pages (PLP), product detail pages (PDP), and throughout the checkout process.

When to use Conversational agent back office

During the Conversational agent integration process, Bloomreach configures the settings for individual agents, such as agent name, logo, language, and tone of voice. You can modify the settings later and perform the following tasks.

Transcripts

• Improve Conversational agent support: Evaluate and review customer transcripts to improve Conversational agent conversation quality. See where Conversational agent sells a product, encourages add to cart, or explains product features.
• Make data-driven decisions: Study transcripts to categorize and analyze customer interactions for informed decision-making.

Agents

• Multi-region support: Launch agents in different languages to support customers from multiple regions, such as the USA and Europe. Note that each agent uses one language. Although agents understand queries in all languages, they respond in their default language.
• Brand agent chats: Tailor agent conversations to your brand’s tone and language.
• Improve response quality: Keep agents in sync with the latest product data for rich and relevant responses.

To modify agents or see transcripts, open the Conversational agent back office.