GITNUXSOFTWARE ADVICE
AI In IndustryTop 10 Best Japan AI Services of 2026
Top 10 Japan Ai Services ranked by accuracy, deployment, and support, comparing Preferred Networks, NEC, and GMO for technical buyers.
How we ranked these tools
Core product claims cross-referenced against official documentation, changelogs, and independent technical reviews.
Analyzed video reviews and hundreds of written evaluations to capture real-world user experiences with each tool.
AI persona simulations modeled how different user types would experience each tool across common use cases and workflows.
Final rankings reviewed and approved by our editorial team with authority to override AI-generated scores based on domain expertise.
Score: Features 40% · Ease 30% · Value 30%
Gitnux may earn a commission through links on this page — this does not influence rankings. Editorial policy
Editor’s top 3 picks
Three quick recommendations before you dive into the full comparison below — each one leads on a different dimension.
NEC
Governed AI workflow provisioning with RBAC and audit logging for model and pipeline changes.
Built for fits when regulated enterprises need governed AI integration with strong schema discipline..
GMO Internet Group
Editor pickGoverned provisioning and administration controls that support RBAC and audit log oriented operations across AI deployments.
Built for fits when IT needs governed AI provisioning, RBAC, and audit trails for managed operations..
NTT DATA
Editor pickRBAC plus audit logging for AI administration and execution traceability.
Built for fits when enterprises need governed AI integration across identity, data models, and automated operations..
Related reading
Comparison Table
This comparison table maps Japan AI services providers across integration depth, data model, automation and API surface, plus admin and governance controls. It highlights practical differences in schema design, provisioning paths, RBAC, audit log coverage, and extensibility for deployment and throughput targets. Readers can compare tradeoffs between enterprise integration patterns and the configuration and sandbox options that affect testing, migration, and operational control.
NEC
enterprise_vendorEnterprise AI services in Japan with systems integration delivery for industrial use cases, including data modeling, operationalization, governance, and API-driven workflow integration.
Governed AI workflow provisioning with RBAC and audit logging for model and pipeline changes.
NEC fits teams that need AI integration with existing back-end systems and clear operational boundaries. Its data model focus supports schema-driven configuration for inputs, outputs, and validation steps, which reduces mapping drift across deployments. Automation and API surface matter for throughput planning since AI inference and workflow steps must match application latency and job scheduling constraints.
A tradeoff appears in the need for upfront integration work to formalize schemas and governance roles before scaling. NEC works well when internal platforms require controlled provisioning, like document processing or predictive maintenance workflows wired into existing enterprise services.
- +RBAC and audit log support governed deployments
- +Schema-driven data model reduces integration mapping drift
- +API and automation hooks fit enterprise system integration
- +Operational monitoring supports run-level troubleshooting
- –Upfront schema and governance setup adds initial effort
- –Workflow configuration can require stronger internal integration ownership
Enterprise platform teams
Deploy AI into internal services
Repeatable provisioning and traceability
Compliance and governance teams
Control AI pipeline changes
Lower governance risk
Show 2 more scenarios
Operations engineering
Run high-volume AI workflows
Higher throughput reliability
Uses monitoring and configuration to manage throughput and failure handling for jobs.
System integration teams
Connect AI to legacy data
Fewer integration regressions
Enforces schema alignment so inputs and outputs stay consistent across systems.
Best for: Fits when regulated enterprises need governed AI integration with strong schema discipline.
More related reading
GMO Internet Group
enterprise_vendorCorporate AI and data services in Japan delivered through group capabilities, supporting industrial data integration, model operations, and managed rollout.
Governed provisioning and administration controls that support RBAC and audit log oriented operations across AI deployments.
GMO Internet Group fits teams building governed AI deployments in corporate environments where infrastructure, routing, and identity controls must stay consistent with existing systems. Integration depth shows up through its administration surface for configuration, resource provisioning, and environment management tied to operational workflows rather than ad hoc experimentation. The data model is oriented around structured service configuration and workload boundaries so schema-driven inputs can be managed across projects and updates. The automation and API surface is geared toward repeatable provisioning and controlled operation, which matters for teams that need throughput planning and deterministic rollouts.
A tradeoff appears when projects require extensive model-level customization that goes beyond service configuration, since the integration focus favors orchestration and governance over deep research-grade parameter tuning. A practical usage situation is migrating an internal AI workflow into a managed deployment that must enforce RBAC boundaries and generate audit logs for operational traceability. Preferred Networks often fits deeper research-to-deployment customization, and NEC often fits enterprise solution delivery patterns, while GMO Internet Group concentrates on integration breadth and admin governance controls for production operations.
- +Strong admin and governance controls for production operations
- +Repeatable provisioning workflows align with enterprise configuration practices
- +API-driven automation supports controlled integration across environments
- –Limited emphasis on model-level customization beyond service orchestration
- –Schema rigidity can slow rapid prototyping without a defined model contract
Enterprise IT operations teams
Provision AI workloads with governed access
Controlled rollout with traceability
Security and compliance teams
Enforce RBAC and audit logging
Auditable usage records
Show 2 more scenarios
Integration and platform teams
Automate AI service onboarding via API
Faster onboarding with governance
Uses automation hooks to provision resources and enforce standard configuration schemas.
Mid-market managed services buyers
Operationalize AI with IT controls
More predictable production behavior
Coordinates deployment steps so infrastructure and identity stay aligned with AI usage needs.
Best for: Fits when IT needs governed AI provisioning, RBAC, and audit trails for managed operations.
NTT DATA
enterprise_vendorSystems integration and industrial AI delivery in Japan that supports end-to-end deployment, integration breadth across enterprise data sources, and operational governance.
RBAC plus audit logging for AI administration and execution traceability.
NTT DATA integration depth shows up in how AI services connect to existing enterprise architectures, including schema mapping from source systems into AI-ready data models. Automation and API surface support hands-on orchestration of pipelines, model lifecycle tasks, and runtime calls that feed downstream apps. Governance controls are designed around RBAC, audit logs, and environment configuration so model execution and administrative actions can be traced.
A key tradeoff is that integration breadth depends on clear access to enterprise data contracts and identity setup for RBAC to be effective. NTT DATA works best when teams have defined target schemas and require controlled provisioning across dev, staging, and production, rather than ad hoc experimentation.
- +Integration-first delivery aligned to enterprise systems and data schemas
- +Governed operations with RBAC and audit log coverage
- +API-oriented automation for repeatable pipeline and runtime orchestration
- –Requires strong data contracts for schema mapping and data readiness
- –Automation surface is easiest to use with dedicated integration ownership
Enterprise architecture teams
Integrating AI into regulated data pipelines
Auditable, schema-consistent AI deployments
MLOps engineering teams
Automating model lifecycle operations
Repeatable releases at higher throughput
Show 2 more scenarios
Customer service ops
Routed AI for contact-center workflows
Controlled automation with traceability
Connects AI inference to case systems and enforces RBAC and audit logging on agent and admin actions.
Compliance and governance teams
Audit-ready AI execution controls
Faster investigations and approvals
Centralizes admin actions and model run events into auditable logs with environment configuration controls.
Best for: Fits when enterprises need governed AI integration across identity, data models, and automated operations.
Accenture Japan
enterprise_vendorIndustrial AI transformation delivery in Japan that emphasizes integration architecture, automation workflows, and governance for enterprise-scale deployment.
Governance-first delivery that maps RBAC, audit log events, and model lifecycle approvals into automated deployment workflows.
Accenture Japan fits into Japan AI services evaluation as an integration-led delivery partner for enterprise deployments. Engagements typically involve end-to-end build and operations work that connects AI services to existing data pipelines, identity, and MLOps workflows.
Integration depth is driven by schema alignment across ingestion, feature stores, and model serving paths, with governance artifacts mapped to enterprise controls. Automation and API surface tend to be defined through project-specific interfaces for provisioning, monitoring, and lifecycle changes rather than a single fixed product API.
- +Project-scoped API integration across data pipelines, model serving, and workflow engines
- +Strong schema mapping work that aligns data model and model input contracts
- +Governance artifacts and RBAC design support audit log and approval workflows
- +Extensibility through custom automation around deployment, monitoring, and rollback
- –API surface and automation breadth often depend on each engagement scope
- –Data model choices can require upfront design time to avoid rework
- –Sandbox and throughput tuning are handled per program rather than standardized
Best for: Fits when enterprises need controlled AI integration, data model alignment, and managed MLOps governance.
Deloitte Japan
enterprise_vendorAI advisory and implementation support for industrial organizations in Japan, including AI governance, controls, and data model design for controlled operations.
Governed AI delivery with RBAC and audit log traceability tied to enterprise provisioning and change control.
Deloitte Japan delivers AI governance and enterprise integration work through consulting delivery plus governed model operations for Japanese enterprises. Integration depth shows up in schema alignment for enterprise data models, plus process mapping for provisioning, RBAC, and audit log workflows.
Deloitte Japan also supports automation and API surface through implementation of AI services that connect to existing systems via documented interfaces and controlled deployment pipelines. Admin and governance controls are reinforced with change management, access controls, and traceability requirements for regulated environments.
- +Strong integration into enterprise data model and schema mapping
- +Defined governance processes for RBAC and audit log traceability
- +Automation focus on deployment pipelines and controlled change workflows
- +Extensibility through integration patterns with existing enterprise systems
- –API surface depends on project scope and chosen AI components
- –High-touch delivery can slow iteration during rapid experimentation
- –Schema and provisioning work adds overhead for narrow use cases
Best for: Fits when regulated teams need controlled AI integration, RBAC, and audit logs across enterprise systems.
PwC Japan
enterprise_vendorIndustrial AI advisory and delivery support in Japan with governance frameworks, audit-oriented controls, and integration planning for production environments.
Governance-led AI program delivery with RBAC-aligned access patterns and audit log oriented traceability.
PwC Japan fits teams that need audited AI program delivery across Japan with consulting governance, architecture, and delivery support. Strength shows up in integration depth with enterprise systems, including schema design for AI data pipelines and joint workshops that map requirements to deployable configurations.
Automation and API surface tend to appear through managed integration work, with emphasis on repeatable provisioning, RBAC-aligned access patterns, and audit log traceability rather than self-serve orchestration. Data model work typically covers end-to-end governance, from labeling and lineage capture to operational controls for ongoing throughput and change management.
- +Integration work aligned to enterprise systems and governance requirements
- +Data model and schema mapping for repeatable AI pipeline delivery
- +Provisioning and RBAC guidance with audit log oriented traceability
- –API and automation surface is driven by service delivery, not self-serve tools
- –Extensibility varies by engagement scope and target operating model
- –Throughput tuning requires PMO-style involvement rather than configuration alone
Best for: Fits when regulated enterprises need end-to-end AI governance plus integration and delivery support.
BearingPoint Japan
enterprise_vendorAI and data consulting for Japan focused on operating models, governance, and integration architecture for industrial AI deployments with controlled rollout.
Governance-focused provisioning with RBAC and audit log alignment across AI-connected enterprise workflows.
BearingPoint Japan differentiates through enterprise integration and governance delivery patterns shaped for regulated Japanese deployments. The service portfolio emphasizes connecting AI outputs into existing enterprise data models, including schema mapping, data lineage, and role-scoped access for operations.
Delivery typically pairs automation and API surface planning with RBAC-aligned administration, audit logging, and provisioning workflows for repeatable rollout. Integration depth tends to be strongest when workloads require coordinated changes across data, orchestration, and controls rather than isolated model demos.
- +Enterprise integration work centered on data model mapping and schema alignment
- +Governance delivery includes RBAC scoping, audit log readiness, and control configurations
- +Automation and rollout planning ties AI outputs to orchestration and operational workflows
- +Extensibility planning supports consistent integration patterns across multiple systems
- –API automation surface depends on client systems and requires upfront integration design
- –Strict governance requirements can slow iteration during early proof-of-concept cycles
- –Data model changes may be required when source schemas do not match target contracts
Best for: Fits when enterprises need controlled AI integrations across data, orchestration, and RBAC-governed operations.
NTT Communications
enterprise_vendorEnterprise managed AI and data service delivery supported by networked infrastructure in Japan, focusing on operational integration and controlled rollout.
Governance-first provisioning with RBAC and audit log coverage for AI workflow access control.
NTT Communications delivers Japan AI Services with a strong integration posture for enterprise environments in Japan. The service focus centers on system connectivity, model deployment governance, and operational controls for production workflows.
Integration depth is supported through documented interfaces for orchestration and data exchange patterns across enterprise systems. Admin and governance tooling emphasizes role-based access controls and auditability for regulated teams managing AI use.
- +Enterprise integration focus for connecting AI workflows to existing Japan systems
- +Governance controls support RBAC and audit log needs for production access
- +API and automation surface for orchestration and provisioning of AI workflows
- +Extensibility for mapping outputs into enterprise data model schemas
- –Deeper integration requires clearer data model mapping work up front
- –Automation coverage depends on target workflow pattern and available endpoints
- –Sandboxing and test throughput controls may lag teams with rapid iteration
Best for: Fits when enterprise teams need managed AI integration with RBAC, audit logs, and production governance.
Frequently Asked Questions About Japan Ai Services
How do Japan AI service providers handle API integration with existing data pipelines?
What integration pattern reduces friction between identity systems and AI access controls?
Which providers are strongest for RBAC and audit log coverage during AI workflow changes?
How do teams migrate an existing AI data model into a governed deployment setup?
What onboarding approach works best when multiple AI workloads must use the same governance controls?
How do these services support extensibility when the required orchestration interface is not fixed?
Where do admin controls matter most for production AI governance?
What common integration failure can occur during schema alignment, and how do providers address it?
Which providers suit regulated environments that need traceability from change control to execution traceability?
Conclusion
After evaluating 8 ai in industry, NEC stands out as our overall top pick — it scored highest across our combined criteria of features, ease of use, and value, which is why it sits at #1 in the rankings above.
Use the comparison table and detailed reviews above to validate the fit against your own requirements before committing to a tool.
Tools reviewed
Primary sources checked during evaluation.
Referenced in the comparison table and product reviews above.
How to Choose the Right Japan Ai Services
This buyer's guide narrows the decision for Japan AI Services providers by focusing on integration depth, data model discipline, automation and API surface, and admin and governance controls. It covers NEC, GMO Internet Group, NTT DATA, Accenture Japan, Deloitte Japan, PwC Japan, BearingPoint Japan, and NTT Communications.
The guidance translates those evaluation points into concrete selection steps for enterprise environments that require governed provisioning, auditability, and operational traceability. The guide also calls out recurring pitfalls seen across consulting and managed integration delivery for regulated rollouts.
Japan AI Services that operationalize AI into governed enterprise data and workflow pipelines
Japan AI Services in this guide cover delivery that connects enterprise data pipelines to managed AI deployments through integration work, schema alignment, and controlled operations. These providers focus on provisioning workflows, runtime orchestration, and governance artifacts such as RBAC and audit logs.
NEC and NTT DATA illustrate this model by combining API-driven system integration with governed operations that support execution traceability and repeatable provisioning. Teams use these services to reduce integration drift, enforce access control, and manage lifecycle changes across multiple AI workloads.
Evaluation signals for Japan AI Services integration, automation, and governance control
Integration depth matters because governed AI deployments fail when data exchange contracts and workflow contracts drift between ingestion, feature preparation, and model execution. NEC and NTT DATA score highly on schema discipline and schema-aligned workflow integration.
Automation and API surface matter because enterprises need provisioning repeatability across environments and controlled lifecycle changes. Accenture Japan and GMO Internet Group emphasize how automation hooks and project interfaces connect into identity, pipeline orchestration, and monitoring so deployments can be managed with admin controls.
Schema-driven data model alignment for repeatable provisioning
NEC centers schema-driven data model discipline to reduce mapping drift during pipeline provisioning and model runs. NTT DATA and Deloitte Japan also emphasize schema and data contract work so multiple AI workloads can be deployed with governed consistency across identity and enterprise data sources.
RBAC plus audit log traceability for admin governance
NEC, GMO Internet Group, NTT DATA, and Accenture Japan all highlight RBAC and audit logging tied to model and pipeline changes. Deloitte Japan, PwC Japan, and BearingPoint Japan extend that governance tie-in into provisioning and change control so access and approvals are traceable end to end.
API and automation hooks for controlled workflow orchestration
NEC provides API-driven workflow integration and automation hooks that fit enterprise system integration. NTT Communications and GMO Internet Group also support API and automation surfaces for orchestration and provisioning so production workflows can be managed with consistent endpoints and environment configuration.
Governed environment configuration and lifecycle change workflows
Accenture Japan and NTT DATA map governance artifacts such as RBAC and audit events into automated deployment workflows for lifecycle approvals. NEC adds operational monitoring for run-level troubleshooting so governance decisions can be validated against execution outcomes.
Operational monitoring for run-level troubleshooting and execution traceability
NEC includes operational monitoring that supports run-level troubleshooting of model execution and pipeline behavior. NTT DATA targets measurable throughput management and repeatable provisioning across multiple workloads, which pairs with execution traceability for controlled operations.
Extensibility through integration patterns tied to client systems
Accenture Japan and BearingPoint Japan design extensibility around integration patterns and custom automation tied to deployment, monitoring, and rollback. Deloitte Japan and PwC Japan focus extensibility through controlled integration patterns rather than self-serve orchestration tooling.
Provider selection framework for governed AI integration in Japan
A structured selection process keeps governance and integration from becoming late-stage project work. NEC and GMO Internet Group are strong examples when the required outcome is governed provisioning with RBAC and audit logs tied to changes.
The decision should start with the integration contracts and administration requirements, then map automation and API surface to the target workflow and environment model. Accenture Japan, PwC Japan, and NTT DATA often fit when the enterprise needs repeatable configuration across identity, data schemas, and operational controls.
Define the data model contract and provisioning schema upfront
Treat the data contract as a provisioning input and require the provider to describe how schema alignment is enforced during provisioning. NEC uses schema discipline to reduce mapping drift, while BearingPoint Japan and NTT DATA emphasize schema mapping and data readiness to avoid rework during rollout.
Require RBAC and audit log coverage for the specific change types that matter
List the lifecycle events that must be auditable such as model version changes, pipeline changes, and access changes. NEC, GMO Internet Group, and NTT DATA explicitly connect RBAC plus audit logging to model and pipeline changes, while Deloitte Japan and PwC Japan tie auditability to enterprise provisioning and change control.
Validate the automation and API surface against the target operational workflow
Confirm how the provider exposes automation and API hooks for provisioning, monitoring, and lifecycle changes into existing orchestration. NEC and NTT Communications focus on API and automation hooks for workflow orchestration and production provisioning, while Accenture Japan often defines automation and API interfaces per engagement scope.
Assess run-level observability and operational monitoring for troubleshooting
Ask for how execution traceability works at the run level and how operational monitoring supports issue isolation. NEC highlights operational monitoring for run-level troubleshooting, and NTT DATA emphasizes governed operations with traceability for AI administration.
Check governance configuration effort and assign integration ownership internally
Plan for upfront governance and schema setup work when the provider uses strict schema and governance discipline. NEC and BearingPoint Japan can require higher initial effort due to schema and governance setup, so internal integration ownership is needed to define and maintain model input contracts.
Match delivery style to rollout speed and sandbox needs
If rapid prototyping and throughput tuning are needed early, scrutinize whether sandbox and test throughput controls are standardized in the provider approach. Accenture Japan and Deloitte Japan handle sandbox and throughput tuning per program rather than standardized tooling, while GMO Internet Group and NTT Communications focus on production operations with controlled provisioning paths.
Which organizations should shortlist which Japan AI Services providers
Japan AI Services are most valuable when the AI use case must plug into governed enterprise systems instead of remaining a standalone prototype. The best provider depends on how strongly the organization needs schema discipline, admin governance controls, and automation and API-driven operational integration.
The segments below map to the best-for profiles used in provider selection for NEC, GMO Internet Group, NTT DATA, Accenture Japan, Deloitte Japan, PwC Japan, BearingPoint Japan, and NTT Communications.
Regulated enterprises needing schema discipline and governed AI integration
NEC is the clearest fit when regulated teams require governed AI integration with strong schema discipline and auditability for model and pipeline changes. NTT DATA also fits when governance controls must cover RBAC and audit log traceability across identity and multiple enterprise data sources.
IT organizations that need governed AI provisioning with RBAC and audit trails for production operations
GMO Internet Group fits when production rollout requires repeatable provisioning workflows aligned to enterprise configuration practices. NTT Communications also fits when managed AI integration must include RBAC, audit logs, and orchestration endpoints for enterprise workflow access control.
Enterprises that need end-to-end integration across identity, data models, and automated operations at scale
NTT DATA supports governed operations with API-oriented connectivity for repeatable pipeline and runtime orchestration. Accenture Japan fits when controlled MLOps governance must map RBAC, audit log events, and lifecycle approvals into automated deployment workflows.
Regulated programs that require governance-led delivery and change-control traceability
Deloitte Japan fits regulated teams that need controlled AI integration with RBAC and audit logs tied to enterprise provisioning and change control. PwC Japan fits when teams need governance-led program delivery with RBAC-aligned access patterns and audit log oriented traceability for ongoing throughput and change management.
Enterprises integrating AI outputs into existing enterprise data models with coordinated rollout controls
BearingPoint Japan fits when AI-connected enterprise workflows require coordinated changes across data models, orchestration, and RBAC-governed administration. It pairs governance-focused provisioning with audit log readiness and rollout planning when multiple enterprise systems must be kept aligned.
Failure modes during Japan AI Services procurement and rollout
Common procurement errors come from under-scoping governance and over-assuming self-serve automation for integration-heavy delivery. Several providers note that schema mapping and governance setup effort is tied to controlled provisioning and auditability.
Another common failure mode is mismatching API and automation surface expectations to the provider delivery style. Providers like Accenture Japan and Deloitte Japan often define automation and API interfaces through engagement scope rather than a standardized product surface.
Treating schema alignment as a one-time mapping task instead of a provisioning contract
NEC and BearingPoint Japan emphasize schema discipline, so buyers should require the data model and schema discipline to be part of provisioning governance. If schema discipline is treated as optional, data contract drift will force rework during pipeline and model input contract changes.
Requesting RBAC and audit logs without enumerating the change events that must be auditable
NEC and NTT DATA tie audit logging to model and pipeline changes, but buyers must specify which lifecycle events require traceability. If only generic access auditing is requested, governance artifacts may not cover the exact approvals and change events needed for regulated operations.
Assuming broad automation and a fixed API surface across all delivery scopes
Accenture Japan and Deloitte Japan describe automation and API surface as project-scoped, so buyers should demand a concrete automation plan for provisioning, monitoring, and lifecycle changes. If integration ownership is not assigned and project interfaces are not clarified early, automation surface gaps emerge during rollout.
Under-allocating internal integration ownership for governance setup and model contract definition
NEC notes that schema and governance setup adds initial effort, and NTT DATA notes schema and data readiness needs strong data contracts. Enterprises that do not assign ownership for integration design and data readiness often see delayed iteration during early provisioning cycles.
Relying on standardized sandbox and throughput tuning when delivery is program-specific
Accenture Japan and Deloitte Japan handle sandbox and throughput tuning per program rather than standardized controls, which can slow early experimentation if throughput is critical. Buyers should ask how test throughput and sandbox controls are configured for the first rollout milestone and not assume uniform defaults across providers.
How We Selected and Ranked These Providers
We evaluated NEC, GMO Internet Group, NTT DATA, Accenture Japan, Deloitte Japan, PwC Japan, BearingPoint Japan, and NTT Communications on capabilities, ease of use, and value with capabilities carrying the most weight, along with ease of use and value each contributing the next largest share. Each provider is scored on how well it supports integration depth, data model discipline, automation and API surface, and admin and governance controls such as RBAC and audit logging.
NEC separated itself by combining very high ease of use with governed AI workflow provisioning that includes RBAC and audit logging tied to model and pipeline changes. That governance-first provisioning lifted both the capabilities score through schema discipline and the operational confidence score through run-level monitoring that supports troubleshooting.
Keep exploring
Comparing two specific tools?
Software Alternatives
See head-to-head software comparisons with feature breakdowns, pricing, and our recommendation for each use case.
Explore software alternatives→In this category
AI In Industry alternatives
See side-by-side comparisons of ai in industry tools and pick the right one for your stack.
Compare ai in industry tools→FOR SOFTWARE VENDORS
Not on this list? Let’s fix that.
Our best-of pages are how many teams discover and compare tools in this space. If you think your product belongs in this lineup, we’d like to hear from you—we’ll walk you through fit and what an editorial entry looks like.
Apply for a ListingWHAT THIS INCLUDES
Where buyers compare
Readers come to these pages to shortlist software—your product shows up in that moment, not in a random sidebar.
Editorial write-up
We describe your product in our own words and check the facts before anything goes live.
On-page brand presence
You appear in the roundup the same way as other tools we cover: name, positioning, and a clear next step for readers who want to learn more.
Kept up to date
We refresh lists on a regular rhythm so the category page stays useful as products and pricing change.