Yearly running cost Assuming 12,000 loads a year
By hand $390K / year 6,000 hours of work
Subscription n.a.
This workflow $281 / year Machine usage only; setup, hosting and review are extra. Do you want to load files or connect a data source? For example, import sales files into warehouse tables and flag rows with missing order references.
One load includes up to 10,000 source rows and 2 source tables.
Load supplied exports Validate approved source data in a warehouse $1,296 – 3,720 setup Connect and schedule sources Keep one approved data flow running $1,872 – 5,400 setup
Which system should it connect to?PostgreSQL warehouse Existing cloud warehouse Write only to agreed warehouse tables. No source-system edits, destructive schema changes or unapproved remapping. A data owner approves mapping changes before scheduled runs use them.
What is included Run a repeatable import from approved exports into isolated warehouse tables.
Import files, check row counts and keep a load history. Connect a source to add scheduled updates and alerts when data is out of date.
Connected source scope: up to 1 source system.
Destination scope: up to 1 warehouse.
Reference application CPU: 4 cores.
Reference application memory: 8 GB.
Standard independent test set: at least 30 loads.
Strict independent test set: at least 60 loads.
What standard do you need? Choose the acceptance targets your provider must demonstrate on your own test material.
Standard Test the defined workload and controls. Correct mapped values: ≥99% Duplicate-free reruns: ≥100% Included Strict Tighter targets and a larger independent test set. Correct mapped values: ≥99.5% Duplicate-free reruns: ≥100% Setup +$360 – 1,200 My own requirements Agree different targets and scope with your provider. Quote separately
How these standards are measured Use independent source snapshots with late records, duplicate keys, schema changes, null values, timezone boundaries and permission failures. Verify reruns and crash recovery do not duplicate destination rows or overwrite source data. Agree the holdout before tuning. Report each metric with its numerator, denominator and exclusions. Count failed and rejected attempts, not just accepted outputs.
Targets for your selected standard What is checked Target Correct mapped valuesCorrect destination values divided by all independently checked mapped values, with omitted values counted as errors. ≥99% Duplicate-free rerunsRerun and interrupted-load cases producing the reference final table divided by all tested rerun and interrupted-load cases. ≥100% Traceable destination recordsDestination records with the correct source and mapping version divided by all destination records tested. ≥100% Approved write boundariesPassed permission and schema-change control cases divided by all tested control cases; any unauthorized write fails acceptance. ≥100%
Proposed targets apply to an independently labeled test set on the agreed workload, not every possible future input. Review each pilot result. Any unauthorized action or sensitive-data exposure fails acceptance.
How quickly do you need each load? Choose a processing-time target for the agreed reference workload; confirm feasibility with your provider.
Within 3 minutes Included Within 2 minutes Setup +$144 – 480 Within 90 seconds Setup +$288 – 840
Timing details Measure from an accepted source snapshot to checked tables, including loading, reconciliation, queueing and permitted retries. Scheduled source availability is separate.
The target applies to at least 95% of agreed test runs, with 1 in progress at a time.
How much should each load cost? Choose a machine-cost cap for the agreed workload. A lower cap needs testing without relaxing the selected quality targets.
Up to $0.30 Included Up to $0.25 Setup +$144 – 480 Up to $0.20 Setup +$288 – 840
Cost details Includes bounded mapping-assistance calls, deterministic load processing, retry allowance and allocated hosting. Source API subscriptions, warehouse storage beyond the reference allowance and staff review are separate.
Reference machine cost: $0.06 – 0.10 per load at 1000 loads a month. The selected cap is a target to test, not a replacement for this estimate.
Where do you want it to run? Run the workflow in your cloud or on your own machine. Choose separately whether it may use an external AI model.
Run it onYour cloud Local environment AI model accessApproved model API No external model calls
Data and access details Run in your controlled cloud environment with scoped access and retained review records.
Use an approved external model with the agreed data-processing terms and minimum necessary input.
Local hosting does not itself keep data offline. Approved business-system connections may still use the network. The baseline uses stable approved mappings; the model is not called once per row. A private-model option changes mapping assistance, not authorization to query the source system. Private running cost must be measured in the pilot.
How do you want to use it? Choose a browser interface, an MCP tool for your current agent, or a standalone agent.
Web page Inspect each load in a private browser interface. Included Your AI agent Call the workflow through MCP; the same access and approval rules apply. Setup +$72 – 360 Standalone agent Use a dedicated agent with the agreed triggers and controls. Setup +$144 – 480