Inside Commerce: Ecommerce Strategy, CX and Technology Podcast
Paul Rogers and James Gurd
Podcast
Episodes
Listen, download, subscribe
EP365: Kibo Commerce Chief Product Officer On Their Unified AI Experience Layer & Why Human-in-the-Loop AI Is The Safer Way To Automate Commerce
Kibo: A unified, model-agnostic approach to agentic commerce For B2B and B2C tl;dr Understand why unified data and one AI interface can outperform a fragmented collection of point agents.See how model-agnostic AI and “bring your own LLM” architecture reduce vendor lock-in and give commerce teams more control over cost and flexibility.Learn how Playbooks automate everyday commerce tasks while preserving human oversight, permissions and accountability. Eric Rosado’s return to Kibo as Chief Product Officer comes at a pivotal moment for ecommerce technology. After previously serving as Kibo’s VP of Product and spending three years as CPO at Monetate, Eric returned to help shape Kibo’s next phase of product innovation, as AI accelerates software development and the way commerce teams operate. The vendor market is evolving rapidly. AI is enabling companies to build faster but that creates new challenges: deciding what to build, keeping pace with constant model innovation and helping customers navigate a crowded market where vendors often promise similar capabilities. Product teams must now design not only for customer experience, but also for “agent experience”, making their platforms easy for AI agents to understand and use. That thinking is central to Kibo AI, Kibo Commerce’s newly available agentic layer. Rather than exposing customers to a growing collection of individual agents, Kibo AI brings nine specialised agents together behind one unified interface spanning commerce and order management. Its unified data model and API-first architecture allow agents to work across catalogue, pricing, promotions, inventory, routing, fulfilment and orchestration without forcing users to understand the underlying complexity. Kibo AI is also model agnostic. Through its “bring your own model” architecture, customers can use OpenAI, Anthropic, Gemini or open-weight models according to their existing technology strategy, performance requirements and budget. This flexibility helps avoid vendor lock-in and gives businesses greater control as model capabilities and token costs continue to change. Our conversation also explores Playbooks: plain-language automations that run on triggers or schedules. They can monitor new products for missing content, flag fulfillment risks, generate reports and connect with external services such as translation engines or Slack. It's important that we emphasise human oversight. AI can analyse, explain and prepare actions, but permissions, audit trails and approval steps remain in place, helping commerce teams automate repetitive work without surrendering control.
Inside Commerce: Ecommerce Strategy, CX and Technology Podcast RSS Feed
