Rantropy
Rantropy provides a verified randomness API and a separate AI action evidence package. The API is delivered as SaaS, while the AI evidence product is supplied through Korea's KONEPS and Venture Nara procurement channels.
The randomness API connects each request and result to a reviewable record. The AI evidence package records AI requests, tool executions, and approvals inside the customer's institutional environment.
For probability items, draws, and random workflows, the API is built to keep the result and the evidence together.
Receipts, hashes, and exports should be useful to customers and auditors, not hidden away as internal logs.
The packaged product records AI requests, tool calls, approvals, and external actions and is available through KONEPS and Venture Nara.
Trust Core
When a system says something happened, a customer or auditor should be able to check the record without guessing. Rantropy builds that checkpoint into the product.
Product Tracks
The verified randomness API is delivered as AWS-hosted SaaS. AI action evidence is a separate packaged product for deployment in institutional environments.
For teams that need to explain how randomness was used in games, events, or API-driven workflows.
A packaged product for institutions that need to preserve and review how an AI system made a decision, called a tool, or triggered an external action.
Rantropy connects API calls, results, timestamps, and validation material so customers and auditors can review them later.
Bind API calls, output hashes, usage, and timestamps into a record that can be checked later.
Connect event order so later edits or missing records become visible. The point is a flow people can inspect.
Do not ask customers to trust the service database blindly. Exported evidence should be checkable on its own.
Package usage, receipts, and validation results in a form that customers, auditors, and procurement reviewers can read.
Use the same randomness features through REST or gRPC to fit the service environment.
Each plan applies its own API key count, quota, rate limit, receipt retention, lookup, and export policies.
Validation
Validation numbers are not the whole story. They should sit where a customer can inspect them, instead of taking over the first impression.
These are internal validation records for the statistical behavior of the randomness engine. Before commercial rollout, they should be packaged with receipts, usage metering, and operating logs that customers can review.
A long-running stream window was captured without reported anomalies.
continuous stream windowEntropy, arithmetic mean, and serial correlation were checked against a stable sample profile.
entropy profileThe captured assessment excerpt recorded no FAIL results.
assessment excerptA representative statistical suite was run in batch form and recorded for review.
sequence batchThe BigCrush battery was run to check the statistical quality of the random generation core.
statistical quality checkRantropy combines randomness APIs, usage controls, and receipt records so product teams can operate and review random workflows in one service.
The current product is an AWS-hosted SaaS service available through AWS Marketplace and direct adoption inquiries. Customer-hosted deployment is not currently offered.
