Same objects. Same data model. Agents on top. No parallel TMS, no fork in the ontology. Just the execution layer OTM was missing.
"There is a need for an entry level solution especially in conjunction and close alignment with OTM."
Bastian Schäffer, DHL Supply Chain
01. The principle
OTM stays where it's strong: planning depth, multi-modal optimization, global trade, financial settlement, ERP integration. Shipsy operates on OTM's objects to execute, then writes back. Three guarantees make the integration true augmentation, not parallel infrastructure.
OTM Order Releases, Shipments, Stops, Service Providers, Equipment. Shipsy reads and writes the same ontology. Custom OTM fields included via schema mapping.
Proven at Expeditors → custom OTM fields mapped through, no re-architecting.
A single normalised store across geographies. Agents trained once, deployed everywhere. Localisation lives in business rules, not in forked data models.
Heineken proved it across 70+ country business units on one data spine.
OTM remains the planning + financial backbone. Shipsy executes against OTM objects, then writes results back. No data divergence, no re-platforming.
Pickup-timeslot booked in OTM ↔ ETA monitored in Shipsy ↔ alert raised, written back.
02. Architecture
The integration architecture for Martin and the OTM team. Same data model, bidirectional sync, no parallel infrastructure. Click through inline, or open the deck full-screen with the controls below it.
Reference architecture from the Heineken implementation. Click to enlarge.
03. The precedent
World's second-largest brewer. 70+ country business units. 2M+ trips a year. Every country had sourced its own trip-management solution. Fractured KPIs, manual POD on paper, manual returnable-packaging tracking, two different stacks for primary vs secondary distribution.
Shipsy replaced the local systems and augmented OTM as the global execution layer. One TMS, one data spine, primary + secondary unified. A case of beer can now be tracked from brewery dock to outlet shelf on a single timeline.
−28%
Sub-contractor payments for excessive stays at customer locations
−50%
Failed deliveries due to "customer not present"
70 → 90%
Fully automated route settlement
+8pp
On-Time Delivery improvement
From the source
Heineken's official video on the primary & secondary distribution transformation. The customer's voice on what changed, why they consolidated onto a single platform, and how local execution is now governed by a single global data spine.
This is what makes the Heineken precedent hard to argue with. It's not a case study deck, it's the customer.
Modules deployed at Heineken
Full Truckload Automation
Load & cost optimization, truck building, vendor indent, execution & visibility
Less-than-Truckload Allocation
Carrier allocation based on performance and cost. Automating hundreds of daily decisions
Returnable Packaging Material
Kegs, crates, pallets. Every movement tracked & reconciled against the outbound shipment
Driver App + GPS + e-POD
End-to-end digital trips. Sign-on-glass, photo, video proof. Paper PODs gone.
Central Control Tower
Single pane of BI + incident management across 70+ countries. Same KPIs, same definitions
What that unlocks for DHL: direct read-across
Slot management
"Customer not present" cut by 50% via automated comms + ETA-aware appointment alerts. Live challenge at DHL → already solved at Heineken scale.
AP-side audit & accessorials
28% drop in excessive-stay sub-contractor payments. Shipsy reduced the stays AND auto-adjudicated the remaining ones. The "massive opportunity" you flagged.
Settlement automation
Route settlement automation lifted from 70% to 90%+. Direct map to your AR-side automated invoicing journey.
Localised → not fragmented
70+ country business units on the same data model. Local rules, global ontology. Exactly the architecture Martin asked about.
04. The value chain
A walkthrough of the AI-native augmentation across the full transport value chain. Same content as the Heineken value-prop workbook, animated end-to-end.
AI-native OTM augmentation: process value chain transformation
05. Capabilities
Capabilities are drawn directly from the Heineken implementation playbook. Real user stories, real acceptance criteria. Each cluster maps to specific OTM objects so augmentation is bidirectional from day one.
Cluster 01
OTM objects:
Equipment · Driver
Live, telematics-aware view of every truck and driver. Self-fleet vs market-fleet differentiation. Hours-of-service-aware ETAs.
Cluster 02
OTM objects:
Order Release · Shipment
Full FTL automation and PTL allocation built around OTM-issued plans. Auto-cascade tendering, sequential trip assignment, override workflows.
Cluster 03
OTM objects:
Stop · Service Provider
End-to-end digital trips. Geofenced delivery + sign-on-glass / photo / video e-POD. Customer comms autonomously close the "customer not present" gap.
Cluster 04
OTM objects:
Equipment · Shipment
Heineken's RPM (kegs, crates, pallets). DHL's analogous: roll cages, dollies, ULDs, totes. Every movement tracked, reconciled, settled against the outbound shipment it belongs to.
Cluster 05
OTM objects:
Shipment · Stop · Event
Single pane of BI across geographies. Same KPIs, same definitions, different localised rules. ETA-vs-OTM-pickup-timeslot alerts route to the right analyst with full context.
Cluster 06
OTM objects:
Voucher · Invoice · Rate
The AP-side play. Excessive-stay adjudication, freight bill audit, accessorial validation, autonomous dispute resolution. Heineken: 70% → 90%+ automated route settlement; $25M+ disputes resolved.
06. AgentFleet
Each agent works inside one operational area. Each agent reads OTM objects, takes action on the execution side, and writes results back. Two of them, Atlas and Nexa, have live walkthrough videos below.
CX Co-worker
Slot booking, WISMO, appointment management, NDR rescue, customer comms across WhatsApp / email / portal.
Heineken: failed-deliveries cut 50%
Ops Co-worker
Planning, sequencing, allocation. Auto-replan on disruption. Hours-of-service-aware ETAs and break-time-aware sequencing.
Heineken: 70% → 90%+ automated settlement
Control Tower
Demo ↓Autonomous incident detection, root-cause routing, auto-remediation. Not a dashboard. A system that acts on OTM Shipment + Stop + Event objects.
Heineken: ETA-vs-OTM-timeslot alerting
Settlement Co-worker
Autonomous carrier & vendor dispute settlement. Reads OTM Voucher + Invoice, validates against contract, settles or escalates.
Heineken: $25M+ disputes resolved
Rating & Audit
Demo ↓Penalty calc, surcharges, accessorial validation, freight bill audit, RCTI generation. The AP-side machinery you flagged.
Heineken: −28% excessive-stay payments
Connectivity Co-worker
Aggregates direct integrations + Project44 + FourKites + internal tools into a single carrier abstraction over OTM Service Provider.
Resolves the connectivity sprawl
Live demo · Atlas
Atlas reads OTM Shipment + Stop + Event objects, detects incidents in flight, routes them to the right analyst with full context. Where rules permit, it remediates without human intervention.
The Heineken ETA-vs-pickup-timeslot alerting pattern, generalised. Threshold per Hub. Backlog per analyst. Segmentable by Hub + Channel.
Open the live Control TowerInteractive demo · DHL-flavoured data · click any incident to walk the workflow
Live demo · Nexa
Nexa pulls OTM Voucher + Rate Record, validates the invoice against contract, classifies accessorials, and routes anomalies to dispute. The AP-side "massive opportunity" rendered as a working pipeline.
Heineken's automated route settlement leapt from 70% to 90%+ with this exact loop running underneath.
Explore the live Control Tower with DHL-flavoured data. Walk any incident from detection to resolution. See agents read OTM objects, route exceptions, and write actions back, in your browser. No setup.
Explore the Control TowerOpens in this tab · Floating "back" pill brings you here · Best on desktop
Detect
Live anomaly & exception alerts
Triage
Routed to right analyst with context
Act
Auto-remediation where rules permit
Audit
Full trail · who · when · what
07. The wedge
You don't need to wait for OTM to land in the Middle East or Southeast Asia. Run a KPI-tied POC in one geography, on the same data model OTM will eventually use.
By the time OTM arrives, Shipsy is already producing outcomes. HCL provides the servicing arm, Shipsy provides the tech, KPIs define success.
POC blueprint
One geography. One pain point.
Pick a country where OTM rollout is years out. Anchor the POC on a single high-impact pain: slot mgmt, AP-side dispute resolution, or e-POD digitisation. Heineken's Brazil pilot is the template.
8–12 weeks to live.
Standard Shipsy deployments: 8–16 weeks. Entry-level scope: 8–12 weeks. The augmentation path skips the months of OTM-side configuration that a full programme requires.
Schema-mapped to OTM from day one.
Shipsy reads/writes Order Release, Shipment, Stop objects. When OTM arrives, the data model is already aligned. No rebuild, no fragmentation.
Joint delivery with HCL.
HCL provides the servicing arm, Shipsy provides the AI-native platform. KPIs are defined upfront and tracked transparently. Both partners aligned on the same outcome metrics.
Precedent · MENA
DHL MENA already proved this shape. Shipsy plugged local 3PLs into DHL's network. COD live in days. Zero new infrastructure.
Precedent · Heineken
Started as a Brazil pilot augmenting OTM. Became the global template across 70+ country business units on a single data spine.
Next
Playground access with live agent demos: Clara, Atlas, Nexa, Vera.
Dummy data, fully interactable for Martin and Bastian. Send a note and we'll provision access within the day.