All projects

Professional project · Spark Universe credit · NDA-protected

Creator event
operations

Design, implementation, administration, and incident handling for a custom, multi-day multiplayer experience broadcast live by a large creator group.

Format
Multi-day custom creator event
Creator load
40+ participating creators
Audience load
10,000+ combined peak viewers
My role
Experience design · development · administration · support
Live duties
Moderation · mediation · incident response
NDA note
Identifying and operational detail removed

Context

One event,
four live layers

The playable build was only one part of the system. Creator comprehension, audience visibility, platform behaviour, and live operations remained coupled throughout each session.

Project contextAction, broadcast, support, and recovery pathsProject structure
L04AudiencePublic surface
Creator channelsCombined live attentionVisible interruption

Every delay, rules question, and recovery could become part of the broadcast.

Broadcast and interpretationPublic feedback and pressure
L03CreatorsParticipant surface
40+ participantsPlay and presentationSupport requests

Creators had to understand the experience while simultaneously performing it for their audiences.

Inputs, rules, and stateQuestions, faults, and conduct signals
L02PlatformPlayable surface
Custom multiplayer experienceDesigned game flowImplemented systems

The build translated the event premise into rules, interactions, pacing, and recoverable state.

Configuration and interventionRuntime state and faults
L01OperationsControl surface
Live administrationCreator supportModeration and incidents

One operating view joined technical state, human context, and the audience-facing consequence.

  1. 01Creator and audience experience were treated as connected outputs, not separate concerns.
  2. 02Design and implementation shared the same working context, shortening the route from intent to live change.
  3. 03Operational decisions considered both a fault itself and the cost of exposing it on stream.
  4. 04Moderation and mediation ran beside technical support because either could block a session.

My role

From brief to
live window

I carried the event from experience definition through implementation and live administration. The sequence shows those hand-offs; exact schedules, partners, and implementation gates remain private.

SequenceExperience production and live administrationNot to scale
  1. 01DefinitionFrame the experiencePremise · player flow · viewer readability · operational constraintsOutput: playable contract
  2. 02ProductionBuild the custom systemsImplementation · configuration · content integration · admin affordancesOutput: event build
  3. 03ReadinessExercise the operating pathsRules clarity · support route · intervention path · fallback conditionsGate: operable under pressure
  4. 04Multi-day live windowRun each sessionAdminister · support · moderate · diagnose · recover · communicateLoop: state → decision → verified return
  5. 05ClosureResolve the remaining stateClose support items · preserve relevant notes · capture operating changesOutput: closed event state
G1 Is the experience legible to a player and a viewer?G2 Can an administrator intervene without creating a larger interruption?G3 Does the recovery route return the affected session to a safe, playable state?

Selected work

Workstream
ownership

My role held creative, technical, and live context across the same event. The table shows the work before and during broadcast without exposing private partner or production detail.

Primary workstreams across preparation and live delivery
WorkstreamBefore liveDuring liveSystem interface
Experience designDefine player flow, rules, pacing, and viewer-readable beats.Interpret deviations against the intended experience.Creators · audience · platform
DevelopmentImplement and integrate the custom experience.Diagnose faults and select safe technical interventions.Platform · administration
AdministrationEstablish operating paths and live controls.Maintain session state and coordinate interventions.Platform · creators
Creator supportPrepare the information needed to enter and understand the experience.Resolve access, rules, and play-blocking questions.Creators · operations
Moderation / mediationSet the route for interpersonal and conduct issues.Contain conflict, mediate where appropriate, protect the event.Creators · operations
Incident responseIdentify failure classes, ownership, and fallback paths.Triage impact, communicate, recover, and verify.All four layers
Operating principle
Optimise for continuity of the live experience, not completion of any one feature in isolation.
Shared context
Creative intent and implementation state stayed in the same decision loop.
Escalation
Technical, creator-support, and conduct issues could be routed without treating them as the same class of problem.
Recovery
A fix was not complete until the affected creator could re-enter a safe, understandable, playable state.

Process

Triage and
recovery loop

Live response joined technical blast radius with human and broadcast impact. The diagram shows the decision logic; real thresholds, communications, and incident details remain private.

Control loopSignal to verified recoveryTechnical and people lanes
I-01 Platform faultI-02 Creator blockedI-03 Rules ambiguityI-04 Conduct conflictI-05 Audience-visible interruption
  1. 01DetectEstablish what happened and who is affected.
  2. 02ClassifyTechnical, support, moderation, or mixed incident.
  3. 03ContainStop the issue expanding across creators or sessions.
  4. 04RouteChoose the technical and/or people response lane.
  5. 05CommunicateGive the affected people a usable next action.
  6. 06VerifyConfirm safe play and understandable state have returned.
Lane ATechnical response

Diagnose state → isolate fault → apply intervention or fallback → verify the playable path.

Lane BPeople response

Clarify context → moderate or mediate → establish next action → verify the safe participation path.

Mixed incidentsRun both lanes under one owner so that a technical recovery does not ignore a human block, or vice versa.
Priority inputsIs play blocked?How many creators are affected?Is the issue public?Is safety or conduct involved?
Contingency scenarios · NDA-safe response classes
SignalContainment moveRecovery conditionPrimary layer
Custom feature misbehavesIsolate the affected path; move to a known fallback where available.Creators can continue without propagating invalid state.Platform
Creator cannot participateSeparate access, state, and rules causes; provide one next action.The creator rejoins or receives an explicit alternative.Creator
Rule is misunderstoodClarify the playable contract without derailing the session.Participants share the same actionable interpretation.Creator / audience
Conduct conflict emergesContain the interaction; move the issue into moderation or mediation.Safe participation can resume, or a clear participation rule is enforced.Operations
Incident is visible on streamPair recovery work with concise, usable communication.The technical or human block is resolved without amplifying confusion.All layers

Outcome

What I
delivered

The event combined a designed and implemented experience with the operating support needed to run it live. Private client material is not reproduced; diagram labels use generic functional terms.

A-01

Experience model

Player flow, rules, pacing, and viewer-facing structure for the custom event.

Internal working material · not reproduced
A-02

Implemented build

Custom systems, configuration, and integrated experience content.

Private implementation
A-03

Operations model

Administration, creator support, escalation, and recovery paths for the live window.

Internal working material · not reproduced
A-04

Incident handling

Technical response, moderation, mediation, communication, and return-to-play decisions.

Private details · logic summarised here
Public outcome

Spark Universe credit; multi-day custom event; 40+ creators; 10,000+ combined peak audience.

Shown here

Layer topology, production sequence, ownership map, triage loop, and contingency classes.

Kept private

Partner requirements, names, communications, exact schedule, source and configuration, internal thresholds, and incident details.