Scenario planning and temporal sandboxes

Simulate high-stakes operational moves. Promote to live with absolute certainty.

Trialling operational decisions in live production is dangerous, and modelling them in disconnected spreadsheets is blind. The JanusKS Scenario Planning Engine introduces high-performance, isolated Temporal Sandboxes. Spin up mathematically complete, zero-copy simulation branches in milliseconds, keeping the full relational depth of your enterprise graph without the storage overhead or multi-hour latency of legacy database cloning. Preview deltas against live operations, run constraint-satisfaction solvers side by side, and publish approved scenarios to production under strict governance and granular access controls.

Two sandbox branches leave the live production line at pointer markers. One is merged back after approval. Sandbox A Sandbox B live graph publish T1 T2
Sandboxes branch from a pointer in the live graph. Only what changes is stored, and only what is approved merges back.
Zero-copy instant branching
Marks the microsecond state of live production, giving full-fidelity simulation environments with zero data redundancy.
Side-by-side live diffs
Real-time visual and mathematical comparison of cost, capacity and timeline trade-offs.
Governed promotion to live
Human-in-the-loop review protocols merge sandbox interventions back into production systems.
Confidential access controls
Role-based permissions keep sensitive reorganisations, bids and downsizing models compartmentalised.

Production is too risky for experiments. Spreadsheets are too brittle for reality.

When an enterprise faces a critical fork in the road, such as bidding on a massive multi-year contract, responding to an acute supply disruption or restructuring team capacity, leadership must test several hypotheses before acting. Traditional tools force an impossible compromise.

  • The copy-paste data nightmare

    Simulation traditionally means cloning entire database instances or exporting massive CSVs into disconnected spreadsheets. By tomorrow, the clone is stale, broken and divorced from live operational reality.

  • No concurrency, no branching

    Enterprise ERPs and scheduling systems are single-track systems of record. They have no concept of parallel operational realities running side by side.

  • Uncontrolled manual updates

    Without governed sandbox-to-live promotion, planners re-key complex schedules into live systems by hand, introducing human error, missing dependencies and costly site clashes.

JanusKS solves this with high-performance, point-in-time pointer branching.

High-throughput zero-copy architecture.

Branches point at the live graph rather than copying it, so each one costs only what it changes.

Live production graph (continuous reality)Live now
Snapshot T1
pointer marker

Sandbox A

Accelerated timeline

Delta tracker

Only modified nodes and edges stored.

Snapshot T2
pointer marker

Sandbox B

Outsourced contingency

Delta tracker

Only modified nodes and edges stored.

Preview and compare engine vs live

  • Cost
  • Margin
  • Fatigue
  • Critical path
Human executive approval

Publish and merge to production

Atomically updates the live graph, then syncs upstream ERPs.

Instant temporal snapshotting with zero data duplication

Traditional cloning duplicates massive databases, consuming infrastructure and taking minutes or hours to initialise. JanusKS uses its bi-temporal graph architecture to deliver full-scale isolation without the wait.

  • Zero data duplication

    Creating a sandbox does not copy your operational data. The engine marks an instant temporal pointer at that exact microsecond in the live graph.

  • Full-fidelity branching

    The sandbox inherits the complete relational depth, history and constraint logic of live operations. Only modified nodes, resequenced tasks or adjusted rates are captured as deltas.

  • Instant multi-branch scale

    Spin up ten, fifty or a hundred concurrent sandboxes instantly, whether your graph holds thousands or billions of nodes.

Previewing and comparing sandboxes to live

Turn subjective operational debates into mathematical trade-off analysis.

The real-time delta engine compares any sandbox branch against live production or against competing branches, automatically computing the structural delta across every dimension.

  • Timeline variance

    Pinpoint which downstream milestones accelerate, compress or slip.

  • Financial exposure

    Surface live cost variances, labour overtime premiums, equipment burn rates and gross margin drift.

  • Constraint and compliance checking

    Instantly verify whether a simulated move violates statutory safety stand-downs, fatigue rules, regulatory hold points or vendor SLA limits.

Multi-branch trade-off matrix

Evaluate competing strategic branches side by side in an interactive decision matrix before presenting recommendations to the board. Illustrative figures.

Live production baseline compared with two sandbox scenarios
Evaluation metricLive production baselineSandbox A: overtime compressionSandbox B: staged subcontracting
Delivery target On schedule 21 days early 5 days late
Total labour cost $3.80M (standard) $4.32M (+13.7% overtime) $3.95M (+3.9% contractor fee)
Gross margin 24.1% 19.8% (margin sacrificed) 23.2% (margin protected)
Resource fatigue risk Normal (72% load) Critical (96% peak load) Safe (68% load)
Regulatory risk Verified compliant Requires council curfew waiver Standard permit envelope
Operational recommendation Active baseline Unviable without client acceleration bonus Recommended operational path

Controlled publishing and merge to live

A sandbox is only as valuable as the certainty with which it can be executed.

Moving from simulation to execution must never happen by accident. JanusKS enforces rigorous human-in-the-loop governance before a sandbox can touch live production.

  • Impact staging review

    Review an itemised checklist of every state change, reallocated resource and adjusted deadline before promotion.

  • Conflict resolution and live drift detection

    If live operations changed while the planner was working, such as an asset breaking down or a permit being delayed, the engine highlights the collision and resolves dependencies before committing.

  • Atomic production merge

    Once authorised, branch deltas merge into the live enterprise graph in one atomic, non-destructive transaction.

  • Closed-loop system sync

    JanusKS automatically broadcasts the new schedule and allocation commitments to operational systems, updating Primavera P6, Procore, Jira or enterprise ERPs through bi-directional webhooks and APIs.

Granular sandbox permissions and confidentiality

High-stakes operational modelling needs strict isolation and compartmentalisation.

Strategic planning often involves sensitive commercial scenarios that cannot be broadcast across the organisation, such as plant consolidations, executive restructuring, confidential M&A or aggressive bidding strategies against competitors.

  • Private and departmental sandboxes

    Create sandboxes accessible only to designated individuals or project leadership, invisible to unauthorised users across both the UI and the API.

  • Agentic permission boundaries

    Autonomous agents working through MCP can be given isolated sandboxes to explore optimisation strategies, with no right to read confidential branches or commit to live operations.

  • Forensic sandbox audit trails

    Every branch created, scenario tweaked and publish event executed is logged in an immutable audit ledger, preserving the institutional memory of why one path was chosen over another.

Role-based sharing and collaboration

Grant granular permissions per sandbox.

Sandbox roles and what each can do
RoleWhat they can do
ViewerInspect simulations, run diffs and review trade-off matrices.
CollaboratorAdjust variables, run constraint solvers and introduce new dependencies.
Approver / publisherHolds the authorised executive credential required to promote a sandbox to live production.

High-impact enterprise use cases

Enterprise use cases for sandboxes and the value of promoting them to live
Operational contextSandbox simulationLive promotion value
Tender bidding and scoping Test uncommitted bids against current capacity without polluting active production schedules. Instantly convert winning tenders into live, fully resourced project plans.
Supply chain disruptions Model multi-week supplier delays and test alternative routing or substitution strategies. Commit re-routed logistics and adjusted delivery milestones across all distribution hubs in one transaction.
Workforce restructuring Confidentially model shift reallocations, department reorganisations or hiring freezes. Promote approved staffing structures directly into workforce management and scheduling tools.
Plant maintenance and turnarounds Simulate accelerated critical path workflows for complex refinery, mine or facility shutdowns. Deploy the optimised maintenance sequence to site teams with no re-keying errors.

Test tomorrow’s decisions without risking today’s operations.

Give your planning teams the freedom to model bold operational strategies, backed by the rigour of deterministic constraint solving and instant zero-copy sandboxing.