1
Activation
One-time discovery, data integration, Knob configuration, evaluation, security work, testing, and launch.
DataKnobs-managed cloud & AIOne managed environment.
One managed environment.
Variable usage stays governed.
DataKnobs can host and operate EKIP, Data Products, data services, and approved AI-model integrations. Fixed platform and operations fees are separated from variable provider consumption, which runs against an approved budget with metering, alerts, and explicit increase policies.
Preapproved usage budgetAlerts and approval gatesHard-stop or degrade policies
Commercial structure
Separate platform value from infrastructure risk.
Managed pricing has four components so customers can see which costs are one-time, fixed recurring, and variable.
2
Platform
Annual DataKnobs / EKIP subscription for governed Data Products, lineage, evaluation, monitoring, and optimization.
3
Managed environment
Fixed annual fee for hosting, operations, observability, security maintenance, backups, and agreed support.
4
Provider consumption
Variable cloud, model, storage, network, and licensed third-party usage funded through an approved prepaid budget.
Managed platform plans
Size platform and operations first; budget consumption separately.
Displayed annual prices combine the DataKnobs platform subscription and managed-environment fee. Activation and variable provider consumption are separate.
Managed Launch
For one focused production Data Product with a controlled workload.
From $60Kper year
$36K platform + $24K managed environment
Scope Launch- 1 production Data Product
- Up to 50 governed Knobs
- Production + non-production
- Managed monitoring and backups
- Standard security baseline
- Usage budget billed separately
Managed Business
For a business unit running multiple related intelligence workflows.
From $156Kper year
$96K platform + $60K managed environment
Choose Business- Up to 3 production Data Products
- Up to 250 governed Knobs
- Managed SDLC environments
- Advanced lineage and evaluation
- Priority operations support
- Usage budget billed separately
Managed Scale
For multiple domains, teams, applications, and higher-volume workloads.
From $405Kper year
$225K platform + $180K managed environment
Contact Sales- Up to 10 production Data Products
- Up to 1,000 governed Knobs
- Multiple managed environments
- SSO/SAML, RBAC, audit support
- Dedicated operations support
- Usage budget billed separately
Managed Enterprise
For regulated, mission-critical, or portfolio-wide managed operations.
Customenterprise agreement
Capacity, security, SLA, and topology based
Contact Sales- Custom Data Product portfolio
- Dedicated or isolated environment
- Custom compliance controls
- Resilience and recovery options
- Enterprise SLA and support
- Committed usage budget
Planning prices shown in USD. Taxes, marketplace fees, custom compliance requirements, bespoke integrations, unusual resilience requirements, and provider consumption are separate unless explicitly included in an order form.
Total commercial model
No unlimited infrastructure promise.
A managed service should be predictable without pretending that highly variable cloud or AI consumption is fixed.
Activation fee
$15K–$150K+
One-time, based on integration and production complexity.
Platform subscription
Annual
Based primarily on production Data Products and governed intelligence.
Managed environment
Annual fixed
Based on topology, environments, security baseline, support, and SLA.
Usage budget
Prepaid variable
Cloud, model endpoints, storage, networking, GPU, and licensed third-party consumption.
Illustrative usage planning
Estimate first. Meter actual usage. Reconcile against evidence.
The bands below are planning ranges only. A workload profile and proof-of-value test should determine the initial production budget.
Controlled
$2K–$5K / month
Illustrative provider-consumption budget for scheduled processing, moderate volumes, selective LLM usage, and standard retention.
Business
$5K–$20K / month
Illustrative provider-consumption budget for several pipelines, frequent evaluation, higher reasoning volume, and multiple data sources.
High scale
$20K+ / month
Custom capacity planning for streaming, multimodal, high-frequency agents, large vector/document workloads, or accelerated compute.
Actual provider cost depends on model and provider pricing, regions, throughput, record/document size, processing frequency, storage, network transfer, retention, availability targets, and architecture.
Financial guardrails
Managed infrastructure should fail safely financially, too.
Recommended protections
- Prepaid monthly or quarterly provider-consumption budget
- Configurable threshold notifications and forecast alerts
- Customer-selected hard stop or graceful-degradation policy
- Explicit approval before budget increases unless pre-authorized
- Workload-level metering and cost attribution
- Daily or agreed-frequency cost forecasting
- Model routing, caching, filtering, and sampling controls where applicable
- Regular capacity and reconciliation review
Avoid ambiguous unlimited promises
- Unlimited LLM calls, tokens, GPU, storage, or data processing
- Silent pay-as-you-go overages
- One fixed price without workload profiling for volatile workloads
- Unbounded third-party API or licensed-data usage
- Capacity increases that happen without a defined approval policy
For very large or unpredictable workloads, customer-owned infrastructure or a separately negotiated committed-capacity model may be a better fit.
Managed vs customer-owned
Choose where the variable-cost risk belongs.
| Consideration | Customer-owned | DataKnobs-managed |
|---|---|---|
| Cloud / model contracts | Customer contracts directly | DataKnobs operates under the agreed managed model |
| Provider consumption | Paid directly by customer | Prepaid and transparently metered |
| Environment operations | Customer or its managed provider | DataKnobs |
| Consumption-control mechanism | Customer budgets and cloud controls | Budget + alerts + gates + stop/degrade policy |
| Best fit | Large, integrated, sensitive, or mature enterprise environments | Qualified teams wanting a turnkey operating model |
| Typical default | Enterprise default | Selective / qualified workloads |
FAQ
Common managed-infrastructure questions.
Does the displayed annual price include cloud and LLM consumption?
No. The displayed managed-plan price combines the DataKnobs platform subscription and fixed managed-environment fee. Variable cloud, model, storage, networking, and third-party consumption uses a separate approved usage budget.
Why not offer unlimited provider usage?
AI and data workloads can vary by orders of magnitude. Separating variable provider consumption avoids hiding risk in the subscription and lets customers choose explicit budgets, alerts, and stop or degrade policies.
What happens when the usage budget reaches its limit?
The operating policy can hard-stop selected workloads, queue or defer noncritical work, switch to lower-cost processing, reduce sampling, or request approval for additional prepaid capacity. The exact policy is agreed before production.
Does DataKnobs mark up provider usage?
The preferred model is transparent provider-consumption reporting plus a clearly stated managed-environment fee. Any taxes, provider commitments, procurement charges, or other pass-through terms should be stated in the order form.
Can we move to our own infrastructure later?
Yes. Portability can be designed into the architecture. Migration services, provider onboarding, data movement, and security integration may be separately scoped while the DataKnobs platform subscription continues.
When should we choose customer-owned infrastructure instead?
Customer-owned infrastructure is usually the better default for very large, unpredictable, deeply integrated, regulated, or contractually constrained workloads, especially when the customer already has strong cloud operations and negotiated provider economics.
Are the usage bands guaranteed prices?
No. They are planning bands only. Actual usage depends on model choice, provider pricing, region, throughput, record or document size, storage, data transfer, retention, availability, and workload architecture.
Start with a measured workload. Scale after the economics are visible.
DataKnobs can profile the workload, estimate a provider-consumption budget, define approval and fallback policies, and launch with explicit financial guardrails.
Request a workload assessment