Government & Federal
Built for the mission side of the firewall.
Use NURA hosted by WE Org, or install it in your environment and assess it as a component under your own RMF process. Deployed in your boundary, all evidence, credentials, and CUI stay inside it.
Two ways to run it.
Hosted by WE Org.
We run NURA and keep it updated, with your data kept separate from other customers. We give your assessors the documentation to review it under your own process.
In your enclave.
Your cloud tenancy, your data center, your Kubernetes. Hardened, signed releases with a software bill of materials on every build.
Air-gap capable.
No phone-home telemetry. No external service dependencies. NURA runs fully disconnected where the mission requires it.
Under your ATO.
Deployed in your environment, NURA is assessed like any other component, with no new cloud service to authorize. We hand your assessors the documentation to do it.
Aligned with where federal compliance is going.
NURA is OSCAL-native and built for the machine-readable, continuously validated model that FedRAMP 20x and RFC-0024 have set for the federal ecosystem. Agencies can use hosted NURA or deploy it in their own environment, both available now.
How a pilot runs.
Scope
Scope 10–15 systems across impact levels and define success criteria with your ISSO team.
Twelve sample systems are scoped across low, moderate and high impact, and success criteria are agreed with the ISSO team.
Deploy
Deploy and configure read-only connectors, load categorizations and tailored baselines.
Read-only connectors attach to each system and baselines load, through week six.
Prove
Live evidence flowing, health scores per system, and your assessors independently verifying the chain and walking an OSCAL package.
Evidence flows and each system’s health fills; the assessors verify the chain independently and an OSCAL package is ready.
You decide
Pilot pricing is fixed and published in the pilot agreement, no open-ended professional-services meter.
At day 90, two equal choices, continue or stop, at a fixed pilot price.
Questions reviewers ask.
Where a capability is scheduled rather than shipped, we say so.
Can we use hosted NURA instead of deploying it ourselves?
Yes. WE Org runs NURA and keeps it updated, with your data kept separate from other customers at the database. Your agency reviews hosted NURA under its own process, and we provide the documentation for that review. If your mission needs everything inside your boundary, or a disconnected deployment, install NURA in your environment instead.
What does “deployed in our boundary” actually require from us?
A Kubernetes environment (or the compute to host one), a PostgreSQL-compatible database tier, and object storage for immutable anchoring, all within your enclave. NURA ships as signed container images with a software bill of materials and a deployment guide. Your evidence, credentials, and CUI never leave infrastructure you control.
Can NURA run air-gapped?
Yes. The platform has no license server, no activation callback, and no phone-home dependency, it is fully functional with zero external connectivity. Updates transfer as signed bundles through your approved media or one-way transfer process, verified offline against public keys pinned at installation. One optional AI-narrative feature is feature-flagged off in disconnected deployments while it completes its move to fully in-boundary operation.
How can our assessors independently verify the chain?
With our software out of the trust path. We provide a standalone verification tool, auditable Python, no platform dependency, that takes any evidence record, retrieves its canonicalized content, recomputes the hash locally, and walks the chain to the anchored root. An SCA methodology guide documents the procedure, a sampling approach, and the canonicalization specification, so the verification can even be re-implemented independently. In demonstrations, we ask your team to pick the record.
How do connectors authenticate, and can they be used for lateral movement?
Connectors use dedicated, least-privilege, read-only service accounts scoped per connector type, with credentials in a secrets vault, never in code or on disk. Agents are outbound-only: they initiate mTLS to the platform and expose no inbound ports, no shell, and no arbitrary execution path, they run only query types defined in a signed task registry. There are no write scopes anywhere in the collection path, and every agent action is itself logged to the evidence chain.
How does NURA handle legacy systems with no APIs?
Four extraction paths, chosen per system: database-level read adapters (read-only SQL), file-drop ingestion (scheduled exports, logs, CSV), host-based agent collection, and, for controls no system can evidence automatically, a structured attestation workflow where a named respondent answers a defined question set and the response is sealed into the same chain with the same integrity guarantees. The target system never needs to know OSCAL exists; OSCAL is generated platform-side.
Can we override the automation: categorization, baselines, tailoring?
Yes, and by design rather than exception. Categorization and baseline outputs are fully overridable, scope-outs, overlays, compensating and inherited controls, organization-defined parameters, with rationale required at the point of decision. Overrides change scope and interpretation; they never touch evidence. Each override is itself sealed to the chain (who, what, when, why), so mission-specific tailoring increases your audit record instead of weakening it.
How does NURA scale, and what are the resource requirements?
The architecture is hybrid, agentless API connectors where targets expose queryable interfaces, lightweight outbound-only agents where host-level presence is needed, on a horizontally scalable Kubernetes platform with a partitioned evidence store. Our honest scale statement: NURA is validated at pilot scale today, and a formal sizing guide with load-test results (resource envelopes per 100 and per 1,000 systems) is a committed deliverable of our pre-deployment engineering program. We won’t assert production scale we haven’t tested, and we’d encourage you to hold every vendor to that.
Who owns the data, and what happens if we leave?
You do, unambiguously, in the agency-deployed model it is your database in your enclave. Everything exports: evidence as canonicalized JSON with hashes and chain metadata, OSCAL artifacts in standard machine-readable form, assessment and POA&M history in open formats. And because integrity is public math, exported evidence remains independently verifiable after you leave the platform. Your audit history’s trustworthiness does not depend on NURA continuing to exist, which we consider the only honest answer to the lock-in question.
A 90-day path to proof.
Ten to fifteen systems. Hosted or in your enclave. Read-only connectors. Live evidence and an OSCAL package your assessors can walk. Then you decide.