← How To

Case Studies · 11 Illustrative Scenarios

Agentic Commerce Case Studies: Trust Architecture in Practice

Principles are only as useful as the situations they survive. These case studies take the AXD frameworks out of the abstract and into the specific - showing how trust architecture, delegation design, and intent alignment play out across financial services, retail, healthcare, travel, and aviation. Each scenario is a composite illustration, not a client engagement, designed to make the discipline concrete.

These case studies are fictional composites designed to illustrate AXD principles. They do not represent any specific organisation or implementation.

Financial Services

01Retail Banking

Autonomous Bill Management in Retail Banking

How a UK high-street bank designed delegation architecture for agent-managed recurring payments

A major UK retail bank sought to enable AI agents to manage customers' recurring bill payments autonomously - identifying better tariffs, switching providers, and optimising payment schedules. The core challenge was not technical capability but trust architecture: how do you design a system where a customer delegates financial authority to an agent that acts when they are not watching, on accounts that contain their salary, mortgage payments, and daily spending? Traditional UX patterns assumed the customer would review every action. The agent needed to act in absence.

Founding Principle 1: Agency Requires Intentional Delegation - the four-tier model ensured every level of autonomy began with an explicit, designed act of delegationFounding Principle 3: Absence is the Primary Use State - the entire system was designed for the customer who is not watching, with return-path narratives as the primary experienceFounding Principle 5: Outcomes Replace Outputs - customers specified desired outcomes (reduce bills by 10%) rather than step-by-step instructions
Read Case Study
02Wealth Management

Agent-Mediated Portfolio Rebalancing

Designing trust calibration for autonomous investment decisions in a private wealth context

A wealth management firm wanted to deploy AI agents capable of rebalancing client portfolios autonomously - responding to market movements, tax-loss harvesting opportunities, and drift from target all…

Founding Principle 2: Trust is the Primary Material - trust calibration governed every aspect of the agent's authority, with progressive exposure replacing blanket permissionsFounding Principle 4: Relationships Have Temporality - the agent's capabilities evolved over time based on accumulated relationship history and demonstrated reliabilityFounding Principle 5: Outcomes Replace Outputs - clients specified portfolio outcomes (growth targets, risk boundaries, ethical constraints) rather than individual trade instructions
Read Case Study
03Business & Commercial Banking

Multi-Agent Treasury Operations

Orchestration visibility and delegation chains in corporate cash management

A commercial bank designed an agentic treasury management system where multiple AI agents operated simultaneously: a cash positioning agent optimising liquidity across accounts, a payments agent manag…

Founding Principle 3: Absence is the Primary Use State - the CFO could not watch four agents simultaneously; the system was designed for a principal who is necessarily absent from most operationsFounding Principle 1: Agency Requires Intentional Delegation - delegation chains made explicit who delegated what authority to which agent, and how authority flowed between agentsFounding Principle 2: Trust is the Primary Material - trust was managed at the orchestration level, not the individual agent level
Read Case Study

Retail & Commerce

04Consumer Electronics

Machine Customer Integration for Consumer Electronics

Designing product information architecture for agent-mediated purchasing decisions

A major consumer electronics retailer discovered that an increasing proportion of product research and comparison was being conducted by AI agents acting on behalf of consumers. These machine customers did not browse product pages - they queried structured data, compared specifications programmatically, and made purchasing recommendations without ever rendering a visual interface. The retailer's entire digital experience was optimised for human eyes: hero images, lifestyle photography, emotional copywriting. None of it was legible to the agents that were increasingly influencing purchasing decisions.

Four Pillars: Signal Clarity - making products machine-readable so agents can discover, evaluate, and recommend without visual parsingFour Pillars: Reputation via Reliability - replacing brand sentiment with verifiable performance data that agents can evaluate programmaticallyFour Pillars: Intent Translation - enabling agents to map human mandates to product specifications accurately
Read Case Study
05Luxury Fashion

Agentic Personal Shopping for Luxury Fashion

Trust architecture for high-value autonomous purchasing in a brand-sensitive context

A luxury fashion house explored enabling AI agents to act as personal shoppers for high-net-worth clients - curating selections, reserving pieces, and completing purchases autonomously. The challenge …

Founding Principle 4: Relationships Have Temporality - the agent's understanding of the client deepened over time, with style memory accumulating relationship history that improved curation qualityFounding Principle 2: Trust is the Primary Material - trust was calibrated through taste alignment rather than transactional metrics, reflecting the unique trust dynamics of luxury relationshipsFounding Principle 5: Outcomes Replace Outputs - clients specified style outcomes ('refresh my evening wardrobe') rather than product specifications
Read Case Study
06Fashion Retail

Agentic Wardrobe Management at Scale

Designing delegation for autonomous clothing replenishment and style coordination

A mid-market fashion retailer designed an agentic wardrobe management system where AI agents could autonomously manage a customer's clothing needs: tracking wear patterns, predicting replacement timin…

Founding Principle 1: Agency Requires Intentional Delegation - the functional/expressive/transitional taxonomy ensured customers explicitly chose what to delegate and what to retainFounding Principle 3: Absence is the Primary Use State - wardrobe management is inherently an absent-state activity; the agent operated continuously while the customer lived their lifeFounding Principle 2: Trust is the Primary Material - returns were treated as trust calibration events, not failures, building a more accurate model over time
Read Case Study

Healthcare

07Patient Care Coordination

Agentic Care Coordination for Chronic Conditions

Designing ethical constraints and failure architecture for autonomous health management

A healthcare provider designed an AI agent system to coordinate ongoing care for patients with chronic conditions - managing medication schedules, monitoring symptom patterns, coordinating between spe…

Founding Principle 1: Agency Requires Intentional Delegation - clinical delegation was explicit, bounded, and revocable, with the clinician as the delegating authorityFounding Principle 2: Trust is the Primary Material - trust architecture in healthcare required zero tolerance for undetected failures, making safety the primary trust signalEthical Constraints Framework - healthcare demanded the most rigorous application of ethical boundaries in any AXD implementation
Read Case Study

Travel

08Travel & Hospitality

Autonomous Travel Planning and Disruption Management

Designing intent architecture and real-time re-delegation for complex multi-leg journeys

A travel platform designed an AI agent capable of planning, booking, and managing complex multi-leg journeys autonomously - flights, hotels, ground transport, dining, and experiences. The unique AXD c…

Founding Principle 5: Outcomes Replace Outputs - intent architecture captured what the traveller wanted to achieve, not the specific bookings they wanted madeFounding Principle 3: Absence is the Primary Use State - disruption management was designed for the traveller who is asleep, in transit, or otherwise unreachableFounding Principle 4: Relationships Have Temporality - preference memory accumulated across trips, improving the agent's understanding of the traveller over time
Read Case Study

Retail & Commerce

10Consumer Electronics - Aftermarket

Autonomous Device Lifecycle Management

Designing agent authority for proactive maintenance, upgrade, and end-of-life decisions

A consumer electronics ecosystem provider designed an AI agent to manage the complete lifecycle of a customer's device portfolio - monitoring performance degradation, scheduling maintenance, negotiati…

Founding Principle 1: Agency Requires Intentional Delegation - per-device delegation ensured customers chose appropriate autonomy levels based on device criticalityFounding Principle 3: Absence is the Primary Use State - proactive maintenance happened before the customer noticed degradation, operating in the background of daily device useFounding Principle 5: Outcomes Replace Outputs - customers specified lifecycle outcomes (maximise device longevity, minimise total cost, prioritise sustainability) rather than individual maintenance actions
Read Case Study

Healthcare

11Pharmaceutical Supply Chain

Agentic Pharmaceutical Inventory and Distribution

Designing compliance-first delegation for autonomous supply chain management in regulated environments

A pharmaceutical distributor designed an AI agent system to manage inventory positioning, demand forecasting, and distribution across a network of hospitals and pharmacies. Pharmaceutical supply chain…

Founding Principle 2: Trust is the Primary Material - regulatory trust required continuous compliance demonstration, not periodic verificationFounding Principle 3: Absence is the Primary Use State - the distribution network operated 24/7 across geographies; human oversight of every decision was physically impossibleEthical Constraints Framework - clinical priority allocation during shortages represented the most demanding ethical constraint in any commercial AXD implementation
Read Case Study

Cross-Cutting Themes

Patterns Across Industries

Across all eleven case studies, several AXD patterns emerge consistently regardless of industry context.

Trust is Always Granular

No case study treated trust as binary. In every scenario, trust was contextual, graduated, and domain-specific. The wealth management client who trusted bond allocation did not trust equity positions. The fashion customer who delegated basics retained expressive choices. Trust architecture must be designed for granularity.

Absence is the Design Condition

Every case study was designed for a principal who is not watching. The retail banking customer asleep when bills are optimised. The traveller mid-flight when connections are cancelled. The patient asleep when medication is managed. Absent-state design is not an edge case - it is the primary use state.

Constraints Enable Autonomy

The most autonomous systems had the most carefully designed boundaries. The pharmaceutical agent that could not violate compliance was more autonomous than one that could. The airline agent operating within mandate parameters made better decisions than one with unlimited authority. Constraints are the architecture of autonomy.

Delegation is a Relationship

Every case study treated delegation as an evolving relationship rather than a one-time permission grant. Authority was graduated, recalibrated, and renewed. The luxury fashion agent earned authority through demonstrated taste. The wealth agent earned equity access through bond performance. Delegation has temporality.

Failure Architecture is Non-Negotiable

Every case study designed for failure as a first-class concern. Healthcare demanded zero silent failures. Banking required automatic rollback. Travel needed cascading disruption management. The quality of failure recovery determined the durability of the human-agent relationship.

Outcomes Over Instructions

Every successful delegation was specified as an outcome, not a task list. 'Reduce my bills by 10%' not 'switch my gas provider'. 'Arrive in Tokyo by Tuesday' not 'book flight BA123'. 'Refresh my evening wardrobe' not 'buy a black dress'. Outcome specification is the language of agentic delegation.

Put Theory into Practice

These case studies illustrate AXD principles in context. To apply them to your own organisation, start with the AXD Readiness Assessment, explore the 12 frameworks in The Practice, or consult the AXD Playbook for a structured implementation guide.