Reference patterns

Operational scenarios where AI systems create durable leverage.

Fifteen patterns across five categories. Each one is a system we've architected before, and an entry point for what we'd architect for you. They're not industry verticals. They’re operational shapes.

UC-MI

Marketing intelligence.

UC-MI-01

Autonomous Content Operations

  1. Brand-governed content drafted, edited, and shipped across every channel.
  2. Editorial standards rise rather than fall as output increases.
  3. Marketing team shifts from production to direction.
UC-MI-02

Campaign Architecture Automation

  1. Campaigns architected from objective to channel mix.
  2. Briefs, assets, sequencing, and measurement plans assembled by the system.
  3. Marketing leaders move from briefing to approving.
UC-MI-03

Brand-Governed Generative Output

  1. Generative output stays inside brand voice by construction.
  2. Tone, taxonomy, and editorial rules enforced at the agent layer.
  3. Brand guardians supervise the system, not every piece of output.
UC-CX

Customer experience.

UC-CX-01

Intelligent Journey Orchestration

  1. One orchestrator runs the customer journey across web, app, support, and email.
  2. The journey behaves like one experience: because one system runs it.
  3. Cross-channel handoffs stop dropping context.
UC-CX-02

Conversational CX Layer

  1. Tier-1 conversations handled with brand-coherent, policy-bound responses.
  2. Escalations to humans arrive with full context attached.
  3. Service capacity scales without proportional headcount increase.
UC-CX-03

Behavioural Optimisation System

  1. The system continuously tests, measures, and optimises journey decisions.
  2. Conversion and retention lift because the system never stops learning.
  3. Marketing science teams move from running tests to designing the test space.
UC-OP

Operational efficiency.

UC-OP-01

Autonomous Workflow Systems

  1. Bounded, rule-bound workflows run without human queue management.
  2. Exceptions surface to humans with the context to resolve them.
  3. Throughput stops being limited by available headcount.
UC-OP-02

AI-Augmented Knowledge Operations

  1. Internal knowledge becomes queryable in plain English.
  2. Policies, runbooks, contracts, and decisions are findable.
  3. Onboarding accelerates; institutional knowledge becomes durable.
UC-OP-03

Cross-System Orchestration

  1. Systems that didn't talk to each other now interoperate via an orchestration layer.
  2. Workflows that crossed three tools collapse into one path.
  3. Integration debt stops compounding.
UC-SD

Strategic decisioning.

UC-SD-01

AI-Augmented Analytics

  1. Data conversations replace dashboard hunts.
  2. Analytics teams shift from chart production to model interpretation.
  3. Decision velocity rises without sacrificing rigour.
UC-SD-02

Strategic Scenario Modelling

  1. Operational futures simulated under named assumptions.
  2. Strategy teams compare scenarios with traceable reasoning.
  3. Boards see the path to a number, not just the number.
UC-SD-03

Frontier Model Decision Support

  1. Frontier models surfaced behind structured prompts and verified data.
  2. Executives use the system to interrogate decisions, not generate them.
  3. Decisioning quality rises without ceding judgement to the machine.
UC-GE

Governance & enablement.

UC-GE-01

Enterprise AI Governance

  1. AI usage is observable, auditable, and policy-bound across the business.
  2. Risk, legal, and security stop discovering shadow AI after the fact.
  3. The Governance & Security Agent has architectural veto authority.
UC-GE-02

Workforce AI Enablement

  1. Every employee operates as a capable operator, not a manual user of point tools.
  2. Capability uplift is structured, measured, and continuous.
  3. Internal AI literacy becomes a workforce-level competency.
UC-GE-03

AI Centre of Excellence

  1. An internal CoE keeps the architecture current after Union Black steps out.
  2. Standards, governance, and capability extension live in-house.
  3. External vendor risk is replaced by an owned operating capability.
Use cases aren't industries. They're operational shapes. The same Customer Conversation Agent solves a problem in fintech, in retail, and in higher education because the operational shape is the same. Union Black · Reference patterns rationale
Recommended next step

Map your scenario to a pattern.

Tell us which pattern most resembles your problem and we'll route you to the right entry engagement: Build Discovery, Strategic Diagnostic, or Architecture Engagement.

Start the conversationRead the methodology