From Chatbot to Agent: KIBO's AI Solution for Shoppers, CSRs, and Merchandisers

Transcript

KIBO’s Gen AI delivers unparalleled agentic commerce throughout the customer and employee journey, surpassing the offerings of most commerce platforms. Let’s talk about what makes KIBO Agentic Commerce unique.

KIBO agents operate across the complete commerce lifecycle, extending far beyond pre purchase engagement.

KIBO’s native vector search and unified order management system equip agents with real time inventory, fulfillment data, and full customer context.

These are crucial differentiators often absent from other commerce solutions.

KIBO’s Agentic architecture is built for tomorrow, leveraging Google’s conversational agents and Vertex AI ready to incorporate any new foundational model.

We will highlight key use cases including order, return, and Wizmo, alongside demonstrations of the customer success agent, the merchandiser agent, and the intuitive KIBO agent admin.

Discover how business users can easily configure and control agents, ensuring these capabilities are governed, composable, and primed for the future. Let’s start with the KIBO shopper agent.

Imagine a customer, Tom, searching for a hiking jacket.

Unlike typical commerce chatbots, KIBO’s agent deeply understands product queries and intent, and instantly checks real time inventory and fulfillment options, because it’s natively connected to KIBO’s OMS PVC.

Tom can ask, Can I get this by Friday? Or Is it available for pickup near me? And get accurate, actionable answers.

When Tom is ready, the agent can place the order and handle payment, all within the chat.

If Tom needs to return an item, the agent guides him through the process, providing pertinent information and generating a return label if required. If he runs into any issues or prefers to speak to a live associate, he can request that the agent transfer him to a customer service rep, who can seamlessly pick up the conversation with full access to the customer’s account, addresses, and order history. This straightforward handling of returns directly mitigates customer churn.

By empowering agents to efficiently manage return requests, Agentic Commerce ensures Tom has a positive experience, even when a purchase falls short of expectations, increasing the likelihood of future purchases.

For Wizmo, Tom simply asks, Is my order ready for pickup? And gets a real time answer. No more searching emails or calling support.

This automation of high frequency post purchase questions drives measurable improvements in CSAT and operational efficiency. KIBO Agentic Commerce is naturally a composable and API driven architecture. Its flat, standalone design allows for seamless integration via API, granting unmatched flexibility without any dependency on other KIBO solutions.

That describes the Storefront Shopper Agent, but imagine extending that same assistance and ease of use to the KIBO UI’s back end.

We are doing exactly that.

Consider Alex, a customer service representative who just fielded a call from Kevin. Kevin placed an order for a windbreaker, but then found the same item for ten percent less elsewhere.

On other platforms, this scenario might trigger time consuming manual order adjustments, potentially requiring the existing order to be voided and a new one created. This could delay pickup, or worse, result in a lost sale.

With KIBO’s Customer Success Agent, the process becomes remarkably simple. The agent can locate and navigate to the order, then apply the necessary appeasement. Kevin receives his product at a fair price, and Alex successfully saves the sale.

The Customer Success Agent offers a comprehensive suite of capabilities, all delivered with the same time saving simplicity. It can manage returns, including issuing return labels, cancel individual items or entire orders, apply coupons, add items to an existing order, and adjust line item quantities. You can even create entirely new offline orders with the Customer Success Agent. This significantly elevates the customer experience, transforms potential challenges into positive outcomes, and ultimately reduces operational costs.

KIBO extends its agentic approach across other areas of the UI, ensuring its business user friendly AI is even simpler to operate. Imagine Taylor, a catalog admin, tasked with creating and modifying products.

Viewing an existing product, such as Feroci pants, or what some would call slacks, Taylor can effortlessly generate product description leveraging KIBO’s merchandiser agent, all based on the product attributes already listed.

If Taylor prompts create a long description of this product and include sizes and colors, the merchandiser agent delivers precisely that. Should Taylor then request make it sound more exciting, the agent provides a more dynamic rendering.

Once Taylor is satisfied with the result, the agent can update the long description, Taylor saves the changes, and swiftly moves to the next product update. This agent is capable of much more. It can also craft new short descriptions, update SEO data, and even translate descriptions based on the regions defined in the parent catalog. This functionality drastically cuts down time to market, enables seamless localization, significantly boosts SEO performance, and enhances both productivity and focus.

KIBO’s Agent Admin serves as your centralized hub for efficiently managing and optimizing all your AI agents, ensuring they effectively assist shoppers, CSRs, and administrators.

Users gain an immediate overview of Agentic statistics from a dashboard, showcasing all available agents. Clicking on a specific agent grants immediate access to quick actions, manage the agent’s playbook, configure its tools, review its conversation history, and thoroughly test its responses and behavior following any modifications.

Playbooks meticulously define agent behavior for each distinct use case, including order management, returns, WISMO inquiries, and catalog interactions. These playbooks can be extensively customized to reflect your brand’s voice and tone, and to establish appropriate escalation paths. Tools operating within these playbooks execute specific tasks such as natural language searches, running a store locator API, or updating payments.

A comprehensive log of all agent conversations is maintained, alongside a dedicated area for thorough agent testing. Robust guardrails are inherently built into the system. Administrators can establish granular data privacy controls to find role based access, implement resource controls, and create stop lists to prevent inappropriate actions.

KIBO’s architecture is both composable and model agnostic, facilitating seamless integration with Gemini, other large language models, or even custom models.

This approach future proofs your AI investment and accelerates the secure deployment of agents. KIBO’s Agent Admin offers centralized agent management, streamlined agent configuration, and enhanced agent performance and testing.

This naturally leads to an optimized customer and employee experience.

KIBO’s Agentic platform leverages Google Conversational Agents and Vertex AI, renowned for their advanced intent understanding and extensibility.

Our model neutral approach allows clients in KIBO to seamlessly integrate new foundational models as they emerge, effectively de risking AI roadmaps.

This same agentic platform forms the bedrock for additional agents currently on KIBO’s immediate roadmap.

This inherent flexibility is paramount as the generative AI landscape continues to evolve, serving as a compelling reason why leading retailers and brands select KIBO for their agentic commerce initiatives.

KIBO’s Gen AI is natively embedded across the entire commerce lifecycle, encompassing order management, returns, WISMO, catalog, and support functions. This empowers business users with advanced AI capabilities, all governed by robust administrative controls, and is architected for future adaptability.