AUTONAR

Security and evaluation

How we make sure it works, and keeps working

Security and testing are part of how we build, from the first release. These are the practices we follow on every project, and you can hold us to them.

Start with a review

Led by an Army veteran. Built by engineers who have shipped production systems for NASA and Cigna, and at Amazon Web Services (AWS), Bloomberg, and Capital One.

Audit log and monitoringEvery action recorded, anomalies flaggedHuman approvalConsequential actions wait for a personSandboxOnly the data and permissions the task needsHarnessContext, tools, and allowed actionsAI model

AI systems

How we make AI systems you can trust

Most AI projects fail on cost, risk, or unclear value rather than on the model itself. We spend our effort on the controls around the model.

Evaluated before launch

We build a test set from your real cases, agree on the accuracy bar before results are in, and keep measuring after launch. A made-up source counts as a failure.

Sandboxed agents

Each agent runs in an isolated environment with only the tools, data, and permissions its task needs. It can't reach anything else.

Human approval where it matters

Low-risk steps run on their own. Anything consequential, like sending, paying, or changing a record, waits for a person. You decide where that line sits.

Audit trail and monitoring

Every question, source, draft, approval, and action is logged. Unusual behavior is flagged, and any agent can be stopped immediately.

Cost limits

Spending caps and cost tracking for each workflow, so running costs are known before you scale.

The full approach is on the AI page: build, evaluate, and govern

All software

How we keep software maintainable

Code review on every change

No change reaches production without a second set of eyes, whether a person or an AI wrote the first draft.

Automated tests

Tests run on every change, so a fix in one place doesn't quietly break another.

Dependency scanning

Every build is checked for known vulnerabilities, and each release ships with a software bill of materials (SBOM).

You own all of it

Source code, infrastructure, documentation, and data belong to you from day one.

Security and data

Your data stays where you put it

We design to the NIST AI Risk Management Framework and to the security controls your authorization (ATO) review will ask for.

YOUR ENVIRONMENTYour cloud account, AWS GovCloud, or your own serversStaffUse the appApplicationRole-based accessAgentSandboxed, scopedModelPrivate or approvedApprovalA person signs offAudit logEvery question, source, draft, approval, and actionPublic internetOutside your controlNothing leaves withoutyour approval

Your environment

Systems run in your cloud account, GovCloud, or on your own servers.

Least privilege

People and agents get the minimum access the job requires.

Encryption

Data is encrypted in transit and at rest, with keys you control.

No training on your data

Your data is never used to train AI models.

Our commitments

What we won't do

These are written into how we work, so you can check them on any project.

×

Ship an AI system without a test set and an agreed accuracy bar.

×

Let an agent take a consequential action without a way for a person to approve or stop it.

×

Use your data to train AI models, or move it outside your environment without your approval.

×

Recommend AI when plain software would do the job better.

Contact

Tell us what you're working on

Send a short note: what the system does today, what isn't working, and any deadlines. An engineer will reply within two business days.

contact@autonar.ai
Email us