Panel Discussion: Engineering Trust Quality in AI-Driven Commerce
summary
Retail teams describe concrete shifts: Super Bowl traffic hits 1.5 million pizzas with about 5,000 orders a minute, peak readiness becomes continuous not a one-time war room, and last season 30–40% bot conversion shows agents already drive revenue. Engineering moves trust into data, contract tests and production-like lower environments, treats agents as customers by validating API/data layers, requires explainability and human-in-the-loop checks, and starts by fixing canonical product data and one solid automated journey.
AI-written summary · about
chapters (46)
- 0:00Start
- 0:05Panel Opens - Always Open: Engineering Trust and Quality in AI-Driven Commerce
- 1:38The Shopper May Now Be an AI Agent, and So May the Tester
- 3:42Super Bowl Sunday: 1.5 Million Pizzas in an Evening
- 5:16"Ready" Doesn't Mean Green Dashboards - It Means Nobody's White-Knuckling
- 6:17Black Friday as the Defining Day for a Fashion Retailer
- 7:50New Look: Peak Running From Black Friday Through Christmas
- 9:23Automation Coverage, Critical Journeys, Performance and Resilience Testing
- 11:26Catalyst Brands: A Multi-Brand Portfolio and Its Many Mini Peaks
- 12:27The Traditional Cycle: Prepare in August, Freeze Code in October
- 13:27The Three-Legged Race: Platform, Operational and Business Readiness
- 15:30Tractor Supply: A Business Built on the Weather Outside
- 17:31The Biggest Pre-Agentic Assumption That No Longer Holds
- 18:31Readiness Is Not a One-Time Event Anymore
- 20:04You Cannot Predict What the Bots Will Do
- 21:05Have You Already Seen Bot Activity? Yes - and Blocked It
- 22:06The Balance: Security vs Not Missing Agent-Driven Sales
- 22:36Twenty Years of E-Commerce and the War Room Era
- 24:06Confidence Now Comes From Production-Like Lower Environments
- 25:38AI Agent Traffic Rose Sharply - What Are You Actually Seeing?
- 26:39The Hard Part: Telling a Bot's Intent Apart From an Attack
- 28:42Are We Ready to Let All Bots Transact? Not Yet. Heading There? Yes
- 30:15The Bouncer Who Now Has to Learn to Host Guests
- 30:45Last Season's Data: 30-40% Conversion Through Bot Traffic
- 31:49Agents Don't Look at Your Store - They Hit the Data Layer
- 33:58When an Agent Can Browse, Compare and Buy - What Changes in Testing?
- 35:32With Decision-Making Bots, It's the Whole Journey That Matters
- 36:03Trust Gets Established When the Agent Proves It Can Be Trusted
- 38:05Layered Security and Trust Deep in the Platform, Not at the Edge
- 39:08Same Price, Inventory and Promotion Whether via ChatGPT or Your Site
- 40:09Agentic Commerce as an Entirely New Revenue Stream
- 42:12Contract Testing Across Consumers and Providers
- 42:43If an Agent Buys for a Customer, You Must Be Able to Explain Why
- 43:44How Test Design Changes With No Single Correct Output
- 45:15The Search Example: "Find Me a Summer Dress Under This Amount"
- 47:47From Search Engine Optimisation to Answer Engine Optimisation
- 50:19The Chipotle Conversational Chat Incident
- 51:49Golden Customer Journeys That Stay Deterministic
- 52:50The Only Reliable Hallucination Check Today Is a Human in the Loop
- 53:20Cash Isn't King Anymore - Context Is King
- 54:22The One Shift to Make First With Agentic Commerce
- 55:56Get Your Data Strategy Right - AI Is Only as Good as Your Data
- 57:27Equip Your Security Team to Make Informed Allow/Block Decisions
- 58:29Keep It Simple: Build One Bot That Does One Thing Well
- 58:59Start With People, Not Technology
- 59:32Pick One Solid Journey and Build Trust Internally From It
related talks
59mPanel Discussion: Strategies for Scaling AI in Quality Engineering
01h 02mPanel Discussion: Embracing Agentic AI: From Autonomous Goals to Enterprise Guarantees
22mAgentic Commerce: The New CX From Discovery to Execution
18mFrom POC to Profit: Deploying Customer-Facing AI Agents That Actually Work
24mPut On Your AI-Colored Glasses
33mEngineering AI Applications: A Conversation with AI Leaders
14mThe Retail Revolution: Real-World Generative AI Applications
47mPanel: AI impact on E-commerce: Friend or a Foe?
