Go back to Market Research
E2open supplies software for supply chain planning, supplier collaboration, logistics execution, international trade, and channel operations. Its e2net network exchanges information between manufacturers, suppliers, distributors, carriers, and other trading partners. Planning applications cover demand forecasting, constrained supply planning, inventory targets, and anticipated shortages. WiseTech Global acquired the company on August 4, 2025. 1 15 16
Overall supply chain score: 4.9/10.
| Dimension | Score |
|---|---|
| Supply chain depth | 5.6/10 |
| Decision and optimization substance | 5.4/10 |
| Product and architecture integrity | 4.2/10 |
| Technical transparency | 5.0/10 |
| Vendor seriousness | 4.2/10 |
E2open’s strongest asset is its position in the flow of operational information and transactions. An ocean booking can be submitted, amended, and confirmed through its documented APIs; demand forecasts can incorporate orders and downstream sales; inventory calculations account for multiple echelons and uncertainty over replenishment lead times. 3 5 17
The planning proposition is less convincing as a system for choosing economically superior actions. Much of the disclosed machinery minimizes inventory against service targets, constructs supply plans from priorities, or improves a forecast that another process turns into decisions. Customer-maintained parameters, reconciliation, and consensus planning remain substantial parts of the operating model. E2open’s own 2026 case study finds that planner adjustments worsened forecast error, yet presents a prospective reduction in planner effort as a productivity gain. Its parent now describes the inherited business as sales-led and says it is establishing a unified product roadmap. Those findings deserve more weight than the expanding vocabulary of cognitive, intelligent, and agentic supply chains. 2 7 14 21
E2open versus Lokad
E2open connects companies and supports the workflows through which they exchange plans, shipments, bookings, and compliance information. Lokad concentrates on computing supply chain decisions: what to buy, produce, allocate, or replenish, and which action offers the best expected economic return given uncertain futures. The distinction matters because a better-connected planning process can still make systematically poor decisions.
E2open’s multi-echelon inventory application minimizes stock subject to service requirements. Its supply planning application uses customer priorities and business rules to construct feasible plans. Lokad’s supply chain perspective starts with the financial consequences of actions, including the asymmetric costs of shortages, excess stock, delays, and capacity consumption. Its inventory optimization approach uses probabilistic forecasts as inputs to decision calculations, rather than treating a consensus forecast or a service-level target as the final specification of business value. 2 3
E2open also addresses uncertain supply. Supply Sensing considers demand, supply, and transportation variability to identify shortage risks. The remaining difference is between identifying a risk and using a coherent economic calculation to select an action across those risks. E2open’s public description supplies shortage probabilities and response options, but leaves the joint uncertainty model and the financial selection rule unspecified. 4
Ocean execution, trade compliance, and the trading-partner network extend beyond Lokad’s offering. The comparison here concerns inventory and production decisions: their economic quality and the recurring work required to maintain the calculations behind them.
A decade of promises and delivery
Forecasting substance preceded the AI umbrella
E2open acquired Terra Technology in March 2016. The announcement identifies an existing business in demand sensing, multi-echelon inventory optimization, and transportation forecasting, with technology developed since 2002. This gives the forecasting portfolio a traceable history beyond recent AI branding. 10
The 2018 forecasting benchmark is unusually informative by enterprise-software standards. It defines forecast error at item-location-week level, distinguishes monthly aggregation for forecast-value-added analysis, and explains its volume-weighted error calculation. It also states that all participating companies use E2open Demand Sensing. This is a study of its customer population, not a neutral competition between forecasting systems. 8
The same year’s Harmony white paper stretched this foundation into a much broader story about cognitive AI and the future supply chain. Its architecture joined data acquisition, decision engines, and a common interface. That structure explains how applications could be brought together; it does not establish that their separate objectives and models became one coherent decision system. 11
Specific releases exposed the maintenance burden
The December 2021 release introduced Supply Sensing and described concrete changes to planning: an in-memory graph engine for business-planning views, a choice of heuristics or cost optimization for supply planning, and simulation controls for more than 80 demand-sensing parameters. 12
The parameter count is particularly revealing. A forecast engine described as self-tuning can still expose a large configuration surface to planners and data scientists. Someone must decide which settings apply, test changes, preserve them through releases, and investigate regressions. The release celebrates control without quantifying that ownership cost. In-memory execution can make these operations faster; it does not remove their necessity or establish better decisions.
Agentic claims now run ahead of measured productivity
An August 2025 explainer presents agents as tireless supply chain coaches that coordinate across the operation and never miss details. A September 2026 announcement describes the Digital Assistant evolving beyond question answering toward coordinated actions and multi-step workflows. The evolution is plausible, but the announcement provides no operating baseline, residual staffing measurement, or financial result for that agentic workflow. 24 13
The historical pattern combines identifiable forecasting and inventory capabilities with progressively broader claims whose delivery is much harder to inspect. The cognitive-AI proposition in 2018 exceeded the integration and specialized decision support described in the accompanying technical documents.
WiseTech’s August 2026 results sharpen the diagnosis. The parent says it is establishing a clear E2open product strategy, aligning teams with product portfolios, and creating a unified roadmap. Its financial report characterizes E2open’s inherited approach as sales-led, with proportionally lower product-development investment, and describes a transition toward a product-led model. These are the acquiring company’s own descriptions of work still underway. 14 15
What the planning machinery actually does
Inventory optimization is primarily a policy calculation
E2open’s MEIO application computes inventory targets across multiple echelons rather than calculating each warehouse independently. It considers network relationships, forecast error, capacity, shelf life, and changing supply parameters. A particularly useful detail is measuring forecast error directly over each product-location replenishment lead time instead of scaling a weekly or monthly error measure mechanically. 3
The stated objective remains minimum inventory for target service levels. A service target encodes a commercial judgment, often very crudely. Two products with the same target can have different margins, substitution opportunities, contractual penalties, and consequences for other purchases. A network-wide stock calculation improves consistency within the specified policy; it does not make that policy economically sound.
Freshness planning illustrates the boundary. The application computes minimum stock for service and maximum stock for a customer-defined freshness target. Its example uses a target under which 99.5% of volume should meet the required remaining shelf life, leaving 0.5% for discounting or disposal. This percentage is a policy assumption, not forecast accuracy. Planners balance freshness, service, production runs, and replenishment; maximum-stock limits can delay production through distribution planning. 9
That is a practical control, but the difficult economic choice has partly been supplied by the user. An economically selected order would compare the expected cost of waste and discounts with lost margin, production economics, and alternatives across uncertain outcomes. A chosen freshness percentage can conceal this trade-off rather than solve it.
Supply uncertainty is recognized; joint decision selection is opaque
Supply Sensing explicitly addresses variability in demand, supply, and transportation, including lead-time effects. It calculates shortage probabilities and supports responses such as alternate sourcing or changes to manufacturing. 4
The public material does not specify how those uncertainties are represented together, how the probabilities are calibrated, or how competing responses are valued. Separate risk indicators can identify likely trouble while leaving a planner to decide whether to expedite, substitute, wait, or accept a shortage. Shared suppliers, capacity, and cash couple these choices; independently attractive interventions may compete for the same resource.
Supply Planning makes the human policy layer explicit. It constructs plans using business rules and priorities for customers, channels, revenue, and commitments, alongside material and capacity constraints. The 2021 release added a choice between heuristics and cost optimization. The disclosed choice of engine does not explain how the modeled costs remain aligned with the changing business. 2 12
The deeper issue is captured by holimization: whether new interactions can enter the decision model economically as the business changes. Adding an alternate supplier or a new customer commitment should not require layers of priorities and exceptions that eventually contradict each other. E2open documents substantial configurability, but supplies little evidence about the cost of keeping those configurations coherent over time.
Forecast accuracy and the value of information
Demand Sensing combines historical demand with newer information such as orders, point-of-sale data, promotions, and channel signals. The 2026 Forecast Engine brief describes a multivariate forecasting layer, automatic method selection, and daily publication of forecasts. It does not disclose the estimators or a reproducible evaluation. 5
A September 2026 consumer-products case reports a 13–38% improvement in MAPE across forecast horizons and up to a 78% improvement in bias. The system covers a seven-week near-term horizon, incorporates orders and downstream data, and refreshes forecasts daily for distribution planning. The case reports two years of historical performance, but omits the absolute error levels, evaluation granularity, and a comparison at matching information cutoffs. 6
That missing cutoff matters. A forecast refreshed after customer orders arrive has information unavailable to an earlier plan. The resulting improvement may justify better information plumbing, but it cannot all be credited to a superior forecasting algorithm. Nor can a signal arriving after a production or procurement commitment repair that earlier decision. The appropriate test follows each signal to an action that was still changeable when the signal became available.
E2open’s 2018 benchmark is more explicit about measurement. Its metric, called MAPE, divides summed absolute errors by summed shipments: a volume-weighted measure rather than an average of individual percentage errors. Forecast error uses item-location-week observations, while forecast-value-added calculations use base-code-location-month observations. Those measures answer different questions and should not be compared as if they had identical granularity. 8
The benchmark also compares safety-stock requirements under three inventory-management scenarios: 14.7 days for traditional single-echelon planning, 12.8 for MEIO using demand-planning error, and 10.2 for MEIO using demand-sensing error. The resulting 31% reduction is a comparison of calculated safety-stock requirements. It does not establish a 31% reduction in customers’ total inventory, or a net financial return after software, integration, and operating costs. The paper’s movement from a modeled stock comparison to bottom-line benefits is too quick. 8
Productivity, configuration maintenance, and accountability
E2open’s September 2026 industrial-manufacturer case offers an unusually useful criticism of manual planning. The statistical baseline had 62% MAPE; the planner consensus had 67%. Planners adjusted 39% of planning buckets. The disclosed process consumed effort while making its chosen forecast metric worse. 7
The case identifies a 30–50% planner-productivity opportunity. Its closing graphic presents the same range as a productivity gain. The body does not establish that the opportunity was realized. Removing harmful forecast edits is only part of the productivity question: the customer still needs replenishment and production decisions, data maintenance, configuration ownership, exception resolution, and monitoring of economic results.
For E2open, several recurring tasks are explicit. The 2021 release exposes more than 80 demand-sensing settings. Inventory planning needs service and freshness policies. Supply planning requires priorities and business rules. The customer security policy assigns customers responsibility for timely, accurate ERP data and reconciliation of ERP information with E2open reports. These responsibilities belong in total ownership cost alongside subscription fees and implementation. 3 9 12 21
Automation must be assessed by the work that disappears. Daily unattended forecasting is useful; faster preparation of a forecast for another consensus meeting leaves the larger workflow intact. An assistant that helps inspect exceptions can increase the throughput of manual planning without making that planning autonomous. E2open’s public cases do not yet quantify the remaining staff effort across the full decision cycle.
There is also a contractual boundary to the marketing. E2open’s published consulting and training terms make fees nonrefundable, provide services and deliverables without the advertised accuracy or fitness warranties, and state that marketing descriptions do not form the basis of the bargain. These terms apply to the specified consulting and training engagements. They leave a buyer needing explicit acceptance criteria, responsibilities, and economic measurements in the statement of work. 23
WiseTech reports $64 million in annualized E2open cost synergies achieved during FY26. That is a measured improvement in the vendor’s cost base. The comparable question for customers—lower ongoing labor and infrastructure cost for equivalent or better decisions—remains unanswered in the reviewed agentic material. 14
Bill of software materials and architectural coherence
E2open’s public subprocessor register gives a more concrete picture than the common-platform language. It identifies different hosting arrangements for specific products, while recruitment material describes distinct engineering requirements. 18 19 20
| Product or evidence scope | Disclosed materials | Architectural implication |
|---|---|---|
| Restricted-party screening and trade automation | AWS locations in Ohio, Ireland, and Germany | Product-specific cloud deployment and regional operations |
| Logistyx | Microsoft Azure in the US and EU | A separate hosting estate within the portfolio |
| China Trade Management | Primary and secondary China Telecom data centers | Another deployment and recovery boundary |
| E2open Digital Assistant | Google Gemini and Vertex AI Search | An external model and retrieval service in the assistant’s processing chain |
| Bangalore software-engineer vacancy | Java, AWS data services including Redshift and Glue; Python and SQL requirements | Data-pipeline engineering with a substantial cloud-service surface |
| Hyderabad software-engineer vacancy | Java, Groovy, Oracle SQL/PLSQL, Protobuf, Python; deployment technologies including Docker, Chef, and Terraform | Another engineering profile with database, serialization, and deployment upkeep |
The hiring entries describe employer-requested engineering skills; the vacancies do not identify the corresponding application. The hosting rows have stronger product attribution. Together they show an estate whose operating costs extend across several platforms and deployment practices.
A common interface cannot eliminate those costs. Separate releases, identity boundaries, data models, monitoring, and recovery procedures require engineering work. Some separation follows geography or product requirements. The commercially relevant question is whether the customer pays for the integration repeatedly through configuration, reconciliation, and services. WiseTech’s continuing work on a unified roadmap makes this a current ownership issue. 14
The public INTTRA API documentation provides a useful example of a clear system boundary. It specifies HTTPS and JSON requests, token-based authentication, booking submission and amendments, carrier schedules, and confirmation webhooks. These interfaces allow transaction workflows to run programmatically without a person repeatedly operating a screen. 17
That capability should receive more weight than conversational access alone. However, E2open’s customer security policy requires prior approval for automated bots operating its applications. Buyers planning their own agents need an agreed route through supported APIs and permissions; unrestricted application automation is not the published default. 21
Security: stated controls and operational responsibility
The customer security policy makes several sensible distinctions: production data should not enter nonproduction environments without written approval, service-account credentials should not be used as personal logins, and customers must maintain appropriate privileges. These are stated operating rules. The page does not explain how the application enforces them or constrains the damage from a compromised integration account. 21
WiseTech’s technical and organisational measures describe authorization controls, encryption, login restrictions, and backups. Their scope is protection of personal data. The document does not establish the isolation model, immutable recovery arrangements, or recovery performance for E2open’s complete operational state. A supply chain can be disrupted by corrupted orders or inventory even when the main incident is not a personal-data disclosure. 22
The Digital Assistant adds a specific external boundary: Google model and search services appear in the E2open subprocessor list. The registry identifies the dependency, but does not describe which operational data reaches the assistant, how retrieved information is permission-filtered, or which actions require authorization. Those mechanisms matter more to the security assessment than certification badges. 18
Supply chain score
Scores evaluate the documented capabilities and operating model. Each dimension is the mean of its five sub-scores; the overall score is the mean of all 25.
Supply chain depth
| Criterion | Score |
|---|---|
| Economic framing | 6/10 |
| Decision end-state | 8/10 |
| Conceptual sharpness on supply chain | 6/10 |
| Freedom from obsolete doctrinal centerpieces | 3/10 |
| Robustness against KPI theater | 5/10 |
Average: 5.6/10.
Economic framing — 6/10. Inventory, waste, transport, and material availability receive explicit treatment. Service and freshness targets still substitute for several financial trade-offs, while the safety-stock benchmark moves too readily from calculated quantities to economic benefits. 3 8 9
Decision end-state — 8/10. The portfolio reaches executable transactions, including bookings and amendments, and supplies inventory targets and constrained plans. This is substantially more operational than a stand-alone forecasting dashboard. 2 17
Conceptual sharpness on supply chain — 6/10. Lead-time-specific forecast error, multiple echelons, shelf-life constraints, and uncertain supply are meaningful details. The relationship between local policies and the economics of the entire system remains underdeveloped. 3 4 9
Freedom from obsolete doctrinal centerpieces — 3/10. Consensus forecasts, service targets, planner overrides, and priority rules remain central. The FVA case demonstrates the cost of manual intervention without establishing a replacement for the larger planning process. 2 7
Robustness against KPI theater — 5/10. The historical benchmark defines its measures and aggregation levels, and the FVA case publishes an unfavorable result for planner intervention. The 2026 demand-sensing case omits absolute error and comparable information cutoffs, while a productivity opportunity becomes a gain in its companion case’s graphic. 6 7 8
Decision and optimization substance
| Criterion | Score |
|---|---|
| Probabilistic modeling depth | 5/10 |
| Distinctive optimization or ML substance | 5/10 |
| Real-world constraint handling | 7/10 |
| Decision automation and labor productivity | 5/10 |
| Resilience under real operational complexity | 5/10 |
Average: 5.4/10.
Probabilistic modeling depth — 5/10. Supply Sensing recognizes uncertainty in supply and transport as well as demand. MEIO uses forecast error over replenishment lead times. Published material stops short of a calibrated joint model connected to economic action selection. 3 4
Distinctive optimization or ML substance — 5/10. Terra’s forecasting lineage, multi-echelon calculations, and the detailed benchmark provide more substance than an AI label. The current forecasting brief leaves methods opaque, and supply planning’s choice of heuristics or cost optimization does not expose the underlying decision model. 5 8 10 12
Real-world constraint handling — 7/10. Material and capacity constraints, alternatives, network relationships, and freshness requirements receive explicit treatment. The remaining weakness is the financial arbitration among competing requirements. 2 3 9
Decision automation and labor productivity — 5/10. Automatic forecasting and transaction APIs remove identifiable work. The most relevant productivity case still reports an opportunity, and the agentic announcements provide no full-cycle staffing measurement. 5 7 13 17
Resilience under real operational complexity — 5/10. Multi-enterprise data and shortage detection support adaptation to disruptions. The accumulated burden of rules, parameters, and customer reconciliation limits confidence that the decision model stays coherent as operations change. 4 12 21
Product and architecture integrity
| Criterion | Score |
|---|---|
| Architectural coherence | 4/10 |
| System-boundary clarity | 6/10 |
| Security seriousness | 3/10 |
| Software parsimony and ownership cost | 3/10 |
| Compatibility with programmatic and agent-assisted operations | 5/10 |
Average: 4.2/10.
Architectural coherence — 4/10. Harmony supplies a common integration and interface story, while the current register identifies separate product hosting estates. WiseTech is still constructing the unified roadmap; public evidence of common decision semantics is weak. 11 14 18
System-boundary clarity — 6/10. INTTRA’s APIs and the product-specific subprocessor register make several execution and hosting boundaries inspectable. Planning-engine interfaces and internal model boundaries are less exposed. 17 18
Security seriousness — 3/10. Published material identifies operating responsibilities and general controls. It offers little detail on enforced tenant isolation, agent permissions, or recovery from corruption of operational data. Those omissions concern the architecture of protection, not the availability of certifications. 18 21 22
Software parsimony and ownership cost — 3/10. Multiple product estates coexist with a large policy surface and customer-owned reconciliation. Neither the configuration burden nor the recurring customer labor is reconciled with the claimed productivity benefits. 12 18 21
Compatibility with programmatic and agent-assisted operations — 5/10. Documented booking APIs provide an actual machine interface. Bot approval requirements and limited disclosure of planning automation constrain the broader agentic proposition. 17 21
Technical transparency
| Criterion | Score |
|---|---|
| Public technical documentation | 5/10 |
| Inspectability without vendor mediation | 5/10 |
| Portability and lock-in visibility | 4/10 |
| Implementation-method transparency | 5/10 |
| Evidence density behind technical claims | 6/10 |
Average: 5.0/10.
Public technical documentation — 5/10. Public API examples, inventory-method details, and a defined benchmark are useful. Equivalent mathematical and operational documentation for the main planning engines remains sparse. 3 8 17
Inspectability without vendor mediation — 5/10. An engineer can inspect booking payloads and authentication flows, while a researcher can examine historical metric definitions. Reproducing a planning result still requires vendor access and undocumented modeling choices. 8 17
Portability and lock-in visibility — 4/10. Transaction APIs expose exchange points. The reviewed documentation does not provide an equivalent export specification for forecasting configuration, supply priorities, or the accumulated inventory-policy model. Those assets affect the cost of leaving. 12 17
Implementation-method transparency — 5/10. Sources disclose signal inputs, forecast refreshes, policy controls, and customer data responsibilities. They give less information about the ongoing effort to test and maintain a configured solution. 5 12 21
Evidence density behind technical claims — 6/10. The dossier contains a longstanding benchmark, concrete parameter disclosures, public API specifications, and cases with actual error figures. Broad agentic claims remain much less substantiated than these narrower capabilities. 7 8 12 13 17
Vendor seriousness
| Criterion | Score |
|---|---|
| Technical seriousness of public communication | 4/10 |
| Technology roadmap integrity | 4/10 |
| Conceptual sharpness | 5/10 |
| Economic accountability | 3/10 |
| Defensibility in an agentic-software world | 5/10 |
Average: 4.2/10.
Technical seriousness of public communication — 4/10. Specific descriptions of forecasting and inventory coexist with claims of agents that never miss details. The latter offers assurance without a defined reliability measure or operating boundary. 3 5 24
Technology roadmap integrity — 4/10. Terra’s methods and subsequent sensing releases show continuity. The leap from 2018 cognitive positioning to 2026 agentic announcements remains weakly tied to delivered outcomes, and the new owner is still unifying the product strategy. 10 11 12 13 14
Conceptual sharpness — 5/10. E2open understands the importance of information across company boundaries and exposes the damage from unnecessary forecast intervention. Its public narrative still conflates better forecasts, more connected workflows, and better economic decisions. 7 8 16
Economic accountability — 3/10. Measured forecast improvements are not accompanied by a complete account of net returns and remaining labor. Published consulting terms distance deliverables from marketing assurances, making contractual acceptance criteria especially important. 6 7 23
Defensibility in an agentic-software world — 5/10. Trading-partner connections, transaction interfaces, and domain-specific information are assets an assistant alone cannot replace. The defensibility of planner-facing orchestration is weaker, especially where the customer’s main benefit remains faster manual review. 16 17 24
Sources
Sources were accessed on November 4, 2026. Product documentation is undated unless a publication or revision date is given.
Supply Chain Planning. Product scope and planning application families.
Supply Planning. Rule-based supply planning, priorities, material and capacity constraints.
[3] Multi-Echelon Inventory Optimization data sheet
Multi-Echelon Inventory Optimization data sheet. Multi-echelon targets, replenishment-lead-time forecast error, and service requirements.
Supply Sensing data sheet. Shortage probabilities using demand, supply, and transportation variability.
[5] Demand Sensing Forecast Engine solution brief
Demand Sensing Forecast Engine solution brief (2026). Forecast inputs, automatic method selection, and daily publication.
[6] CPG manufacturer demand-sensing case study
CPG manufacturer demand-sensing case study (2026-09-29). Seven-week demand-sensing deployment and reported forecast improvements; anonymous customer.
[7] Industrial manufacturer forecast-value-add case study
Industrial manufacturer forecast-value-add case study (2026-09-28). Baseline versus planner MAPE, share of manually adjusted buckets, and prospective productivity range.
[8] 2018 Forecasting and Inventory Benchmark Study
2018 Forecasting and Inventory Benchmark Study (2018). Metric definitions, aggregation levels, study population, and modeled safety-stock comparison.
[9] Freshness inventory planning solution brief
Freshness inventory planning solution brief. Freshness targets, minimum and maximum inventory, and production-planning interaction.
[10] E2open Acquires Terra Technology
E2open Acquires Terra Technology (2016-03-07). Terra acquisition and the origins of its sensing and inventory applications.
[11] Building a Digital Supply Chain
Building a Digital Supply Chain (2018). Historical Harmony architecture and cognitive-AI positioning.
[12] Quarterly technology update: Supply Sensing and Global Logistics Orchestration
Quarterly technology update: Supply Sensing and Global Logistics Orchestration (2021-12-01). Original release PDF documenting Supply Sensing, supply-planning engine options, and more than 80 parameters.
[13] Control Tower Technology Value Matrix announcement
Control Tower Technology Value Matrix announcement (2026-09-15). Vendor announcement describing the Digital Assistant and agentic workflow direction.
[14] WiseTech Global FY26 results presentation
WiseTech Global FY26 results presentation (2026-08-26). Unified-roadmap work, E2open cost synergies, and the inherited sales-led model.
[15] WiseTech Global FY26 financial report
WiseTech Global FY26 financial report (2026-08-26). Audited acquisition date and financial reporting context.
[16] e2net multi-enterprise network
e2net multi-enterprise network. Trading-partner connectivity and the network proposition.
[17] INTTRA Ocean Execution API v1 specification
INTTRA Ocean Execution API v1 specification. Public JSON specification behind apidocs.inttra.com: booking methods, authentication, and webhooks.
[18] WiseTech Global group subprocessors: e2open-specific register
WiseTech Global group subprocessors: e2open-specific register. Product-specific hosting and the Digital Assistant model/search providers.
[19] e2open software engineer vacancy, Bangalore
e2open software engineer vacancy, Bangalore. Employer-authored vacancy describing data-platform engineering requirements; publication date unavailable.
[20] e2open software engineer vacancy, Hyderabad
e2open software engineer vacancy, Hyderabad. Employer-authored vacancy describing language, database, and deployment requirements; publication date unavailable.
[21] e2open Customer Security Policy
e2open Customer Security Policy (2024-10-04). Customer data reconciliation, privileges, service accounts, and prior approval for bots.
[22] WiseTech Global Technical and Organisational Measures
WiseTech Global Technical and Organisational Measures (2024-11-18). Published personal-data protection measures, including authorization and backup controls.
[23] e2open Consulting and Training Terms and Conditions
e2open Consulting and Training Terms and Conditions. Terms for consulting and training, including warranty limitations and marketing descriptions.
[24] Work Smarter: Agentic AI Sets a New Standard for Supply Chain Mastery
Work Smarter: Agentic AI Sets a New Standard for Supply Chain Mastery (2025-08-06). Vendor explanation of agentic AI and its claimed operating properties.