Xenith Sort
The reasoning layer above parcel sortation control.
Xenith Sort sits above warehouse control and dimensioning systems for CEP hubs and 3PLs. It ingests scan, chute, DWS, WMS, and PLC signals; runs no-read recovery, revenue leakage checks, and an ops copilot with tool-grounded answers. Physical and billing actions route through human approval. Shadow Mode logs recommendations before any write path.
Integrate-don't-replace over WCS/MFC · Temporal or inline approval core · synthetic demo metrics labeled in workshops
Deployed as a scoped engagement for your stack - not a self-serve trial or live sandbox.
Why teams look for this
Problems this pattern addresses
No-reads pile into manual re-sort
Typically 3–5% of parcels fall to exception lanes. Control rooms see dashboards but lack confidence-scored next actions.
Silent DWS vs WMS revenue leakage
Dimensioning and billed weight diverge. Carriers can leave 2–4% on the table until month-end reconciliation.
Dashboards without a governed write path
Ops copilots that invent divert commands fail security review. Recommendations need RBAC and audit before Assist.
Where it sits in your stack
How it works
No-Read Recovery
Exception queue, OCR hooks, re-induct proposals with citations.
Revenue Leakage Engine
DWS × WMS × rate cards → billing correction proposals only.
Ops Copilot
Tool-grounded Q&A over live hub signals; it cannot invent PLC writes.
Approval Inbox
Operator / supervisor / admin RBAC with append-only audit.
Governed AI
Safety, audit, and rollout rules
Read the Governed AI doctrine for how Xenqube separates model output from money and safety actions.
Zero-LLM-Trust on money and motion
Language models summarize and draft. Deterministic engines and humans own billing corrections and physical divert.
Shadow before Assist
Connect → Shadow → Assist → Scale. No WCS write path until supervisors trust reject rates.
Proof and measurement
Workshop and shadow-trial metrics are labeled when synthetic. Live SLAs require a named pilot agreement.
| Metric | Target / demo | Notes |
|---|---|---|
| Leakage flagged (demo) | €1.2K+ MTD | Synthetic demo ledger, labelled in workshops |
| Exception diagnosis (target) | ≥80% faster | Shadow-trial success criterion, not live SLA |
| WCS writes in PoC | None | Shadow Mode only until Assist phase |
Who this is for
Use cases
Rollout
Phased trust model
Read-path adapters for scan, chute, DWS, WMS signals.
Log recommendations; measure operator reject rates.
Approved writes only; dual-control on billing corrections.
Multi-hub rollout with shared audit and playbooks.
What's included
Typical engagement deliverables
Scoped to your stack and compliance needs. We deploy the working system into your environment and hand over runbooks - not a sandbox login.
Signal map from your WCS/DWS/WMS read paths
Shadow-mode recommendation log (no writes)
Approval workflow wired to your RBAC model
Synthetic demo KPIs labeled for workshop review
Questions buyers ask
Frequently asked questions
- Does Xenith Sort replace our WCS or fusion suite?
- No. It sits above read paths and proposes governed actions. Rip-and-replace is not the default pitch.
- Can it write to MFC on day one?
- Not in Shadow Mode. Assist phase opens approved writes after you accept reject-rate and audit evidence.
- Are demo € figures real customer results?
- Workshop demos use synthetic labeled metrics unless a named pilot is commercially approved for publication.
Related products
More in Industrial AI