How agentic orchestration works
Our methodology in six diagrams: how we research, how the Agentic Hub coordinates agents and people, how work moves through a run, and how we decide what agents may do on their own.
- 01ResearchNarrow questions, evidence first
- 02OrchestrateThe Agentic Hub coordinates agents
- 03GovernAutonomy earned, guardrails in code
- 04MeasureGates, budgets, and feedback
- What we measure shapes the next research question.
The research loop
How every pattern we publish or build into Marketix moves from a narrow question to evidence.
- 01QuestionA narrow question that evidence can answer
- 02MapThe real workflow, data, and people
- 03PrototypeThe smallest system that can answer it
- 04EvaluateTest sets, rubrics, and shadow runs
- 05BuildTurn it into a product capability
- 06PublishFrameworks and findings, shared openly
- LogEvidence logEvery step is recorded. Each loop ends in a decision: continue, build, or stop.
- Then back to step 01 with the next question.
Agentic Hub architecture
The coordination layer between people, specialist agents, and the systems they act on.
- People
- Goals and plansOutcomes, priorities, budgets
- Guardrails and policyLimits, claims, consent rules
- Approvals and reviewConsequential work, escalations
- Agentic HubCoordinates every agent and person against one plan
- PlannerGoals into tasks
- RouterRight specialist
- Shared contextPlans, facts, audiences
- ApprovalsStates and review
- Run logEvery decision
- Specialist agents
- Research
- Content
- Demand
- Lifecycle
- Analytics
- ConnectorsTool connectorsScoped permissions, hard guardrails, every call logged
- Systems
- CRM
- CMS and site
- Ads
- Analytics
- Knowledge base
- DataSignal data layerIdentity, consent, intent, and the unified event stream
- Outcomes and learning feed back into the hub's plans.
Anatomy of an orchestration run
How one piece of work moves through the hub, from trigger to learning.
- Tools and channels01TriggerAn event or request starts the run
- Agentic Hub02PlanTasks, owners, and risks
- People03Review planPeople adjust and approve
- Agentic Hub04DelegateRoute each task with full context
- Specialist agents05ExecuteParallel where possibleEscalates to people when unsure
- Agentic Hub06CheckGuardrails and consistency
- People07ApproveConsequential work
- Tools and channels08ActPublish, send, or update
- Agentic Hub09LearnMeasure and log results
- Results inform the next plan, back at step 02.
The autonomy ladder
Five levels of agent independence, and the evidence needed to climb them.
- Level 0ManualPeople do the workFull human control
- Level 1AssistAgent drafts and researchesPeople choose what to use
- Level 2Approve to actAgent prepares actionsPeople approve each one; guardrails in code
- Level 3Act and notifyAgent acts within limitsPeople can reverse; guardrails in code
- Level 4AutonomousAgent acts within guardrailsPeople review samples; guardrails in code
- Promote one level at a time, with evidence. Step down when the failure budget is spent.
Evaluation gates
The stages an agent passes before it touches live work, and the criteria at each gate.
- Stage 1Offline evalsGolden dataset and rubric
- Gate: meets quality bar on test set
- Stage 2Red teamDeliberate failure hunting
- Gate: severe failures fixed
- Stage 3Shadow modeReal inputs, no real actions
- Gate: matches human baseline
- Stage 4Limited rolloutAssist or approve to act
- Gate: within failure budget
- Stage 5ProductionMonitoring and failure budget
- New failures in production become new test cases at stage 1.
Signal data flow
How data moves from sources to agents and channels, and how outcomes flow back.
- Sources
- Website
- Product usage
- CRM
- Email and ads
- Events
- Support
- Signal data layer
- IdentityLinked conservatively
- ConsentPreferences, opt-outs
- IntentMeaningful patterns
- Event streamOne timeline
- Consent gate: only eligible data passes
- HubAgentic HubDecides what to do with each signal, within guardrails
- Agents and channels
- Sales alerts
- Lifecycle
- Paid media
- Content
- ResultOutcomesMeetings, pipeline, retention
- Attribution and results feed back into the data layer.
Four ideas behind every diagram
People set goals and limits
Agents pursue goals that people define, inside guardrails that people write and code enforces.
One hub coordinates
Plans, shared facts, routing, and approvals live in one place, so many agents behave like one system.
Everything is recorded
Every run, decision, and outcome is logged, which is what makes evaluation, audit, and learning possible.
Autonomy is earned
Agents gain independence one level at a time, with evidence, and step back down when the evidence changes.
See the architecture in the product
Marketix is built on these diagrams: specialist agents, governed connectors, and a hub that plans, routes, and records every run. Sign up for early access.
