Contact us

Review of Centric Software, Product Lifecycle and Retail Planning Software Vendor

By Léon Levinas-Ménard

Last updated: November 4, 2026

Go back to Market Research

Centric Software supplies product lifecycle management, retail planning and pricing, market intelligence, visual collaboration, and product experience management software. Its applications cover product specifications, sourcing and costing, assortment and merchandise planning, allocation, replenishment, markdowns, and product content. The company is a Dassault Systèmes subsidiary serving fashion, retail, and other consumer-goods businesses. Its current Planning & Pricing offering combines capabilities developed through the Armonica Retail and aifora acquisitions. 1 2 7 9

Overall supply chain score: 4.4/10

Dimension Score
Supply chain depth 4.8/10
Decision and optimization substance 4.4/10
Product and architecture integrity 4.2/10
Technical transparency 4.0/10
Vendor seriousness 4.4/10

Centric addresses economically consequential retail decisions, particularly the interaction between prices, markdown timing, and inventory. Customer evidence includes substantial reductions in markdown work and improved margins. Its weakness is the distance between those demonstrated applications and the broader claim of a unified AI decision engine. Public material describes connected data, business rules, forecasts, and workflows far more clearly than a joint economic optimization of purchasing, pricing, allocation, and uncertain supply.

The acquisition history also matters. Planning, pricing, market intelligence, and product content arrived through different products. Centric publishes concrete integration claims, and current hiring identifies a combined Planning & Pricing team. Yet the technical record still shows several application and database environments, along with significant implementation and production-support work. Their coordination contributes to total ownership cost. 5 19 20

The September 2026 Centric AI announcement adds a companion, MCP connectivity, and task agents. Generating a bill of materials or routing an approval can save time, but these activities do not establish autonomous supply-chain optimization. The new AI offering should be evaluated separately from older planning and pricing results that now appear alongside its marketing. 15 16

Centric Software versus Lokad

Centric’s retail scope includes an important economic lever: changing the price changes demand. Its pricing offering considers initial prices, promotions, markdowns, and inventory movement across the product lifecycle. Its PLM integration brings product costs and supplier information into planning. These are substantive connections between commercial and operational decisions. 3 5

Lokad’s inventory optimization service selects operational decisions by their expected financial contribution under uncertainty. Purchases, transfers, allocations, and production orders compete for scarce resources through a common economic perspective. Supply Chain Scientists maintain the logic that automates recurring decisions. Introduction to Supply Chain develops the underlying design perspective.

Centric’s public implementation instead emphasizes a connected environment for merchandising, pricing, product development, and planning teams. Business users set strategies and rules, inspect recommendations, and coordinate decisions. Automating one specialized task does not establish automation of the complete buying and inventory process. 3

The consequential comparison is whether the system makes the economic trade-offs consistently. A markdown can accelerate sales while destroying more margin than a transfer would cost; a transfer can be pointless when demand will disappear before arrival. Shared screens and synchronized data make these choices easier to discuss. A decision engine must also value the uncertain outcomes and select the action.

What is actually optimized?

Pricing has economic substance, but the objective remains only partly visible

Centric describes lifecycle pricing, elasticity modeling, allocation, replenishment, and stock rebalancing. The January 2025 release adds what-if comparisons between strategies such as maximizing profit and reducing end-of-season stock. It also describes a separate forecasting service and a common taxonomy for combining internal information with competitor product and pricing data. 3 4

This is more specific than a generic claim to improve forecast accuracy. Price is a decision variable, not just an input. Inventory and markdown timing are economically coupled: waiting preserves a higher selling price but risks a larger later reduction or unsold stock.

The public documentation does not explain how the system estimates the causal response to a price change. Observed sales at a lower price may reflect promotion, seasonality, product age, availability, or a changing customer mix. A model that predicts those historical observations well can still recommend the wrong future price. Elasticity estimation needs validation of the intervention, not merely a fit to past sales.

The joint scope is also unresolved. Centric discusses alternative profit and inventory strategies and lets users guide the system with business rules. It does not publish an objective or reproducible example showing how it reconciles price, transfer costs, uncertain arrival times, substitutions, returns, and constrained buying budgets. Price automation can therefore leave consequential purchasing and allocation trade-offs outside its calculation.

Lead-time awareness is not a probabilistic supply model

Centric’s replenishment material recognizes shelf and distribution-center capacity, minimum orders, packing requirements, and short or long lead times. Its 2024 AI/ML article also mentions supplier performance and variations in supply lead times. These considerations belong in a retail decision model. 6 26

The missing element is how uncertainty enters the selected action. A delayed shipment, a demand peak, and an approaching end-of-season date can interact to make a previously attractive replenishment worthless. A point estimate of each factor, or a manually chosen adverse scenario, does not give their joint probability or expected economic consequence.

The published materials do not expose calibrated distributions for either demand or future lead times, nor a method that carries those distributions through price and inventory choices. More frequent updates can improve visibility after a supplier changes a commitment. They do not replace anticipation of the possible changes before the business commits money.

Forecast claims lack the measurement needed to assess them

The January 2025 announcement promises greater forecast accuracy and up to 25% faster performance. The performance figure concerns the software improvements described in the release; it is not a 25% forecast-accuracy improvement. No forecast metric, comparison protocol, horizon, or out-of-sample results accompany the accuracy claim. 4

Centric’s forecasting article does identify the relevant granularity: each SKU, channel, and location. That is the right level of concern for local inventory decisions. However, a discussion of weather, trends, competitors, and AI is not evidence that the forecasts are calibrated at that granularity. The public record does not establish how stockouts, new products, sparse observations, and changes in price are handled in forecast evaluation. 25

Forecast error also remains an intermediate measure. A better average forecast can coexist with worse purchases when the costly tail risks, supplier delays, or margins are represented poorly. The economic result should be evaluated on the resulting prices and orders, not inferred from the forecasting label.

Customer results: substantial gains with specific scopes

Leder & Schuh: measurable markdown automation

Leder & Schuh reports markdown work falling from five people for five days to two people for 1.5 days: 25 person-days became three, an 88% reduction. Coverage spanned nine countries and 12,000 SKUs. Reported financial gains exclude promotions: price reductions fell 2.5%, while a stated 1% margin gain generated €3 million on €300 million turnover. That calculation corresponds to one percentage point of turnover. 12

Implementation began in 2019; the case still describes replenishment as newly initiated and Centric Planning as prospective. The result therefore supports a specialized pricing deployment. Software fees, support, and policy maintenance are absent from the return calculation, leaving its all-in economics unmeasured.

Guess: distinguish company performance from software attribution

Centric’s Guess case reports planning cycles falling from one month to one week and describes better coordination of countries, channels, and assortments. It also explains that Centric and Guess configured an allocation module during the project, removing a step previously performed in another application. These are concrete workflow improvements. 13

The financial claim requires precision. The case’s operating-margin movement from 5.6% to 12% corresponds to Guess’s adjusted, non-GAAP operating margin in fiscal 2020 and fiscal 2022. Guess’s own results release confirms those figures; reported GAAP operating margin in fiscal 2022 was 11.8%. The issuer attributes the improvement to higher initial markups, lower markdowns, and lower occupancy costs, among other effects. 14

The customer describes Centric as enabling its strategy. That is a credible contribution claim, not a controlled estimate of software-generated profit. The case itself discusses changes in brand positioning, product costs, operations, and real-estate economics. The current product page’s banner of 100% greater operating margins strips away much of that context. 3 13

The deployment helped Guess execute a broader strategy; the software’s independent return and repeatability remain unmeasured.

Technology roadmap: acquired intelligence and a new agent layer

Date Public development What the evidence establishes
February 2019 aifora announces NKD’s adoption of intelligent price automation, with continuously learning algorithms and progressively greater automation. The current pricing lineage had a pre-LLM AI narrative and named customer testing. The release gives no quantitative test protocol or realized ROI. 11
November 2021 Centric acquires Armonica Retail. Merchandise financial planning, assortment, allocation, replenishment, and forecasting broaden the earlier PLM perimeter. 7
November 2022 Centric acquires StyleSage and promotes AI-powered product matching and competitive intelligence. The claim concerns matching, benchmarking, and market information; it does not establish autonomous supply-chain decisions. 8
September 2023 Centric acquires aifora and promises integration with the existing suite. Pricing and inventory automation arrive through an acquired product with prior deployments. 9
January 2025 Centric releases the Pricing & Inventory improvements discussed above. Functional development continues after acquisition. 4
February 2025 Centric signs an agreement to acquire Contentserv for €220 million enterprise value. Product information and experience management extend the portfolio into another application domain. 10
September 2026 Centric introduces a common AI offering with a companion, MCP, task agents, and AI applications. Document and workflow automation expands; the disclosed examples do not demonstrate joint probabilistic optimization. 15

The early AI claims need to be assessed at their actual scope. A 2019 pricing model and a 2022 image-matching system can perform bounded tasks without the general capabilities associated with later LLMs. The unsupported step is converting those labels and acquisitions into evidence of broad intelligent decision-making. The old aifora release promises optimal prices and continuously improving predictions, yet publishes neither the assumptions behind optimality nor a quantified validation.

Current documentation describes synchronization between planning, product costs, and supplier information. The October 2026 architect role also identifies a combined Planning & Pricing organization. These are concrete integration signals. Shared data and a shared team do not establish one economic decision model. 5 19

The 2026 AI narrative creates an attribution problem of its own. The AI product page places earlier customer stories alongside new claims about agents and AI-driven returns. The Guess pandemic deployment predates the new platform’s launch. It supports the older planning product’s results, rather than demonstrating that the newly introduced agents produced them. The page acknowledges that outcomes combine software, AI, and human judgment, but does not separate their contributions. 13 16

Bill of software materials

Current, dated employer postings reveal more of the implementation than the product brochures.

Product or operational scope Attributable components Consequence
Retail Planning and Planning & Pricing implementations The October 2026 technical architect role specifies SQL Server, Microsoft Azure, APIs, and ETL integration. Customer deployments require database design, integration blueprints, sizing, configuration, upgrades, and performance work. 19
Broader application engineering A January 2026 role specifies Java, SQL, TypeScript, REST services, Docker, and Kubernetes, including custom-action features. The application estate includes substantial executable business logic and container operations. 22
Production database support An April 2026 DBA role explicitly covers customer PostgreSQL deployments, slow queries, locking, vacuum behavior, configuration, and availability. Production performance requires specialist diagnosis and maintenance across hosted and customer-managed environments. 21
Portfolio DevOps An August 2026 role names Kubernetes, Rancher, Docker, Helm, and Terraform, and lists troubleshooting of Kafka, RabbitMQ, PostgreSQL, SQL Server, Oracle, and MongoDB. The portfolio carries several infrastructure and data technologies, with corresponding upgrade, integration, and incident-handling responsibilities. 20
aifora data ingestion Theobald’s customer account describes Xtract Universal moving SAP data into an Amazon S3 data lake and an ETL pipeline. This documented deployment used daily transfers and partitioned extraction to avoid timeouts. 24
Generative-AI delivery An August 2026 role covers retrieval, agents, tools, evaluation, guardrails, interfaces, and production readiness. Agent development adds model evaluation, tool permissions, and release validation to the engineering workload. 23

SQL Server and Azure belong to the retail-planning implementation role; the broader DevOps list covers portfolio operations. Database tuning, cross-system integration, container operation, and several product lineages impose real costs. The public record does not quantify how much is eliminated by the current combined offering or passed through to customers as implementation and support effort.

Configuration maintenance and total ownership cost

Centric’s technical architect role is explicit about work after the sales demonstration: installation, configuration, upgrades, performance optimization, and integration with ERP systems and data warehouses. The PostgreSQL role adds query tuning, configuration changes, recovery, and customer escalations. These responsibilities make the recurring engineering burden visible. 19 21

There is also economic configuration. Leder & Schuh sets business rules and can choose a more aggressive or defensive markdown posture. Those choices encode the trade-off between inventory clearance and margin. If the market changes, someone must decide whether the rule, demand response, price schedule, or stock policy should change. The software’s ability to recalculate quickly does not determine which assumption needs revision. 12

This is the issue emphasized by Holimization: the mathematical model itself must evolve. A useful implementation makes it economical to revise objectives, constraints, and decision logic as the business learns. A collection of configurable modules can instead distribute those assumptions across rules, ETL transformations, dashboards, and specialized services.

The Theobald integration case provides a small but concrete example. Initial SAP extraction encountered timeouts and required splitting data into blocks with vendor assistance. The resulting daily feed supported the customer successfully. The case shows both that integration problems can be resolved and that apparently straightforward data movement has a continuing engineering surface. 24

Neither major customer story supplies a full lifetime cost calculation. Markdown labor savings and shorter planning cycles are valuable, but the denominator should include policy maintenance, data engineering, support, and implementation specialists as well as end users.

Agents, productive work, and accountability

The September 2026 release claims 90–98% time savings on selected manual tasks, 2–5% gross-margin improvement, and gains of up to 5% in sell-through. The named task example converts a sketch or supplier document into a bill of materials and technical pack. Those activities can remove clerical work, but the release does not provide a named cohort, baseline, measurement period, or calculation of total planning labor behind the figures. 15

Centric’s current AI page distinguishes routine tasks from human judgment. It says uncertain details should be flagged, and its FAQ leaves people accountable for reviewing and confirming decisions. That is a meaningful product boundary. It also limits the autonomy claim: an agent that prepares a document or proposes a next action is not yet selecting and executing the recurring economic decisions of the supply chain. 16

MCP connectivity adds little evidence of decision quality. It exposes existing tools and context to external AI systems. Relative to building a reliable economic policy, such an adapter is a low-effort addition. Its prominence in the platform announcement is a warning that familiar workflows may be receiving a new interface while the underlying decision problem remains unchanged.

Productivity should be measured through specific work removed, scope expanded, and outcomes maintained or improved. Savings in a document-entry task cannot be extrapolated to the entire planning function.

Economic accountability remains incomplete. Public materials attribute improvements to Centric-enabled decisions while retaining human confirmation and responsibility. The cases do not describe a systematic review of recommendations that were followed yet reduced profit, or how software defects are distinguished from poor customer policy. Without that feedback, positive results can validate the platform while negative results remain attributable to user judgment.

Security: assess the defaults beneath the certifications

Centric’s public security page describes company-device restrictions, encrypted backups, centralized logs, dependency checks, and testing before release. These address concrete parts of the operating environment. The 2025 system description also discusses named access roles, network filtering, VPN access, and interfaces for data import and export. 17 18

One documented default is weak. In the PLM data workflow described on printed page 13 of the 2025 SOC 3 report, encryption at rest is available but enabled only on customer request. That makes a basic protection dependent on customer configuration. The report covers calendar 2025; the public material reviewed does not establish that this default has since been removed.

The encryption passage describes the PLM workflow. The broader product portfolio also requires explicit boundaries between customer data, administrator privileges, integrations, and agent tool access.

Centric’s AI materials promise permission control and human oversight without exposing the detailed authorization and execution boundaries of the MCP and agent layer. The current public evidence is insufficient to evaluate how a compromised connector or mistaken tool invocation is contained. Security certifications and release checklists do not answer that architectural question.

Supply chain score

The 25 criteria carry equal weight. Each dimension is the arithmetic average of its five sub-scores.

Supply chain depth

Criterion Score
Economic framing 6/10
Decision end-state 7/10
Conceptual sharpness on supply chain 4/10
Freedom from obsolete doctrinal centerpieces 3/10
Robustness against KPI theater 4/10

Average: 4.8/10.

Economic framing — 6/10. Lifecycle pricing links demand, margin, markdown timing, and inventory. Customer evidence includes a monetary result. The public decision criterion does not extend clearly to joint purchasing, allocation, supply risk, and constrained capital. 3 12

Decision end-state — 7/10. Pricing, markdown, allocation, replenishment, and stock-transfer outputs affect operations directly. The markdown case demonstrates meaningful execution automation. Other parts of the suite remain centered on plans, scenarios, and team coordination. 4 6

Conceptual sharpness on supply chain — 4/10. Centric recognizes that prices influence demand and that supply feasibility matters. Forecast precision and connected information nevertheless carry much of the public argument, without a corresponding account of uncertainty and economic selection. 5 25

Freedom from obsolete doctrinal centerpieces — 3/10. Merchandise financial targets, planning cycles, scenario review, and configurable rules remain prominent. Integration improves these processes but does not establish that their targets and local objectives align with overall economic returns. 3 12

Robustness against KPI theater — 4/10. Detailed customer numbers support real gains. Broad margin banners and new AI metrics compress important differences in scope, attribution, and timing, while forecast claims lack a published evaluation protocol. 3 4 14 15

Decision and optimization substance

Criterion Score
Probabilistic modeling depth 2/10
Distinctive optimization or ML substance 4/10
Real-world constraint handling 6/10
Decision automation and labor productivity 6/10
Resilience under real operational complexity 4/10

Average: 4.4/10.

Probabilistic modeling depth — 2/10. Demand variation, supplier performance, and lead times appear in the material. Calibrated outcome distributions, their dependencies, and joint propagation into decisions are not disclosed. Forecasting and scenario simulation alone do not establish this capability. 25 26

Distinctive optimization or ML substance — 4/10. Pricing automation, product matching, and forecasting have identifiable product histories. Public descriptions provide limited algorithms, benchmarks, or causal validation, leaving the distinctive mathematical contribution difficult to assess. 8 9 11

Real-world constraint handling — 6/10. Retail capacity, lead times, packaging, assortment, supplier information, and country-specific pricing are recognized. Their simultaneous treatment under uncertainty and economic priorities is less clear than their individual coverage. 5 6

Decision automation and labor productivity — 6/10. The markdown process has a quantified before-and-after workload, and Guess reports substantially shorter planning cycles. New agent claims lack comparable measurement of all work, including review, exceptions, and maintenance. 12 13 15

Resilience under real operational complexity — 4/10. Named deployments cover multiple markets and channels. Public evidence does not validate the unlimited-SKU language or show how newly discovered economic interactions enter a tested revision of the decision logic. More rules and integrations can increase maintenance without repairing the underlying objective. 6 19

Product and architecture integrity

Criterion Score
Architectural coherence 4/10
System-boundary clarity 5/10
Security seriousness 4/10
Software parsimony and ownership cost 3/10
Compatibility with programmatic and agent-assisted operations 5/10

Average: 4.2/10.

Architectural coherence — 4/10. Centric publishes real integration developments and now organizes planning and pricing together. The portfolio still spans several product lineages and technology environments, with extensive integration and support responsibilities. 4 19 20

System-boundary clarity — 5/10. Product origins, some data exchanges, and the distinction between task agents and human decisions are visible. The boundaries between a shared data foundation and shared economic optimization remain insufficiently specified. 7 9 16 24

Security seriousness — 4/10. Public material describes actual access and backup controls. The 2025 PLM encryption default and limited disclosure of agent containment prevent a stronger assessment of security by construction. 17 18

Software parsimony and ownership cost — 3/10. Current operations roles expose a broad technology estate and specialist database support. Customer-specific rules and integration work add costs beyond hosting and licenses; public case studies do not provide the resulting lifetime total. 19 20 21

Compatibility with programmatic and agent-assisted operations — 5/10. REST interfaces, batch integrations, automated pricing, and the newer MCP and task agents provide real automation surfaces. Their existence does not yet establish a coherent, economically autonomous planning program. 15 18

Technical transparency

Criterion Score
Public technical documentation 4/10
Inspectability without vendor mediation 3/10
Portability and lock-in visibility 3/10
Implementation-method transparency 5/10
Evidence density behind technical claims 5/10

Average: 4.0/10.

Public technical documentation — 4/10. Product pages, release descriptions, employer postings, and a system report expose functions and operating components. Public algorithm specifications, forecasting protocols, and developer-level integration documentation remain limited. 4 18 22

Inspectability without vendor mediation — 3/10. Readers can evaluate stated scope and customer outcomes. They cannot reproduce the pricing policy, inspect its uncertainty model, or validate the claims of joint optimization on a published example. 3 4 6 12

Portability and lock-in visibility — 3/10. Imports, exports, and interfaces are described, but there is no public demonstration of transferring configured economic logic, business rules, histories, and dependent applications into an independently operable replacement. 18 24

Implementation-method transparency — 5/10. Current roles and customer cases reveal integration design, configuration, custom modules, and production support. The effort required after initial deployment is visible qualitatively but poorly quantified. 13 19 21

Evidence density behind technical claims — 5/10. The markdown case has detailed workload and financial measures, and Guess’s financial figures can be checked against issuer reporting. Broader AI and forecasting claims have much less precise validation. 12 14 15

Vendor seriousness

Criterion Score
Technical seriousness of public communication 4/10
Technology roadmap integrity 4/10
Conceptual sharpness 4/10
Economic accountability 5/10
Defensibility in an agentic-software world 5/10

Average: 4.4/10.

Technical seriousness of public communication — 4/10. Functional releases and concrete customer accounts coexist with unbounded scalability language, unmeasured forecasting superlatives, and broadly attributed AI benefits. The latter claims exceed their published validation. 4 6 15

Technology roadmap integrity — 4/10. Acquired capabilities have identifiable histories and some visible integration follow-through. Reusing older product outcomes in the newer AI narrative obscures which innovation produced which benefit. 4 13 16

Conceptual sharpness — 4/10. The link between pricing and inventory is economically sound. A unified data and workflow environment is nevertheless presented more clearly than a unified decision criterion, particularly when supply outcomes are uncertain. 3 4 5 6 26

Economic accountability — 5/10. Customer reports include actual margin and labor improvements, not just adoption metrics. Full cost accounting, balanced analysis of unsuccessful recommendations, and the vendor’s share of responsibility remain unclear. 12 13 16

Defensibility in an agentic-software world — 5/10. Retail-specific pricing automation and substantial product data can support valuable recurring decisions. MCP, conversational access, and document generation provide weaker differentiation; their value must be measured beyond accelerating familiar human workflows. 12 15 16 23

Source dossier

All sources were accessed on November 4, 2026. Hiring dates come from the employer’s public Workday records. Historical descriptions and customer results retain their product and period scope.

[1] Company history and portfolio

Centric Software. Published: Undated. Current portfolio, subsidiary status, and acquisition timeline.

[2] Centric PLM product scope

Centric Software. Published: Undated. PLM scope, including product development, specifications, sourcing, and costing.

[3] Centric Planning and Pricing

Centric Software. Published: Undated. Current combined offering, pricing and planning functions, customer quotations, and headline performance claims.

[4] Pricing and Inventory performance release

Centric Software. Published: 2025-01-21 (release body). Functional changes, performance claim, forecast-service integration, strategy comparisons, and shared taxonomy. The release body is dated January 21; the page header says January 20.

[5] Connecting PLM and Planning

Centric Software. Published: 2025-12-25. Current description of synchronization between planning and live product, cost, supplier, and lead-time information.

[6] AI inventory replenishment

Centric Software. Published: 2023-11-14. Retail constraints, allocation and replenishment scope, automation claims, and unlimited-SKU language.

[7] Armonica Retail acquisition

Centric Software. Published: 2021-11-30. Acquisition announcement and the acquired planning product’s functional perimeter.

[8] StyleSage acquisition

Centric Software. Published: 2022-11-15 (release body). Acquired matching and competitive-intelligence capabilities. The body is dated November 15; the header says November 14.

[9] aifora acquisition

Centric Software. Published: 2023-09-14 (release body). Acquired pricing and inventory product, prior customers, and planned integration. The body is dated September 14; the header says September 13.

[10] Contentserv acquisition agreement

Centric Software. Published: 2025-02-25. Agreement announcement and enterprise value, together with the product-information and experience-management scope.

[11] NKD adopts aifora price automation

aifora. Published: 2019-02-27. Historical vendor release describing customer tests, pricing strategy choices, and intended automation; expected benefits are not treated as realized financial results.

[12] Leder and Schuh markdown automation case

Centric Software. Published: Undated; describes implementation in 2019. Vendor-hosted customer account with markdown workload, margin calculation, policy choices, and the scope of subsequent projects.

[13] Guess planning case

Centric Software. Published: Undated; describes pandemic-period deployment. Vendor-hosted customer account covering planning cycles, allocation configuration, strategic changes, and reported operating margins.

[14] Guess fiscal 2022 results

Guess, filed with the SEC. Published: 2022-03-16. Issuer’s SEC-filed results confirm adjusted margins and identify several drivers of the company’s performance.

[15] Centric AI introduction at NRF Europe

Centric Software. Published: 2026-09-10. New companion, MCP, task-agent and application offering, with task-time and financial claims.

[16] Centric AI product and FAQ

Centric Software. Published: Undated. Current product boundaries, human confirmation and accountability, and presentation of customer outcomes.

[17] Security and Data Privacy

Centric Software. Published: Undated. Current public security statements, access restrictions, backups, and release checks.

[18] 2025 SOC 3 system description

Centric Software. Published: 2025 reporting period. System description for January–December 2025. Printed page 13, PDF page 16, describes customer-requested encryption at rest in the PLM data workflow.

[19] Technical Architect, Retail Planning and Pricing, JR428

Centric Software on Workday. Published: 2026-10-28. Current employer posting linking retail-planning implementation to SQL Server, Azure, integration, configuration, upgrades, and performance work.

[20] Senior DevOps Engineer, JR347

Centric Software on Workday. Published: 2026-08-31. Employer description of portfolio infrastructure and database troubleshooting responsibilities; alternatives and customer systems are not treated as one installed stack.

[21] Production PostgreSQL DBA, JR192

Centric Software on Workday. Published: 2026-04-28. Production database tuning, availability, incident response, and support across customer hosting environments.

[22] Senior Software Engineer, JR16

Centric Software on Workday. Published: 2026-01-06. Application development responsibilities naming Java, TypeScript, SQL, containers, and custom actions.

[23] Senior Frontier AI Deployment Engineer, JR344

Centric Software on Workday. Published: 2026-08-21. Generative-AI delivery responsibilities and preferred experience, distinguished from proof of shipped product capabilities.

[24] aifora SAP integration with Xtract Universal

Theobald Software. Published: Undated. Integration supplier’s account, including the aifora CTO’s description of SAP extraction, S3, daily delivery, and timeout mitigation. The review relies on the aifora-specific portion.

[25] AI demand forecasting in retail

Centric Software. Published: 2024-04-10. SKU/channel/location forecasting claims and discussion of inputs; no evaluation protocol accompanies them.

[26] AI and ML across the retail cycle

Centric Software. Published: 2024-06-27. Public account of feature extraction, clustering, elasticity, supplier lead-time considerations, and AI across the retail lifecycle.