We only ship
what you can check.
Before a Nexgile digital worker reaches production, its workflow is readable, its behaviour can be tested on sample data, and — if you choose — it is attacked with the techniques adversaries use. This page is how we prove it — and what we haven’t earned the right to claim yet.
EU data protection
California privacy
In transit & at rest
Oversight & approvals
Security & availability
Top of the roadmap — target date: being finalized.
Information security
AI management system
Accessibility conformance
“In place” reflects controls and protections active today. “Roadmap” marks formal third-party certifications we are actively pursuing — we do not claim a certification before it is awarded. Contact security@nexgile.com for current status and documentation.
You’ll notice some chips say Roadmap. That’s deliberate. We publish what’s true, not what’s aspirational — and every claim on this site is labeled the way we label these. Top of the roadmap: SOC 2 Type II — target date: being finalized. While certifications are in flight, here’s what we offer today instead: self-hosted deployment inside your existing compliance boundary, the audit-trail specification, and sanitized export artifacts your security team can review on request.
Security, engineered into every delivery.
The software Nexgile generates ships with security built in — and our own platform is operated to enterprise standards.
- Encryption in transit and at rest
- Identity and access controls with least-privilege by default
- A STRIDE security threat model generated with every export
- Continuous monitoring, logging and a defined incident-response process
- Responsible-disclosure program for reporting vulnerabilities
Your data stays yours.
Nx·Studio is a design-time studio: workers can be tested on sample data generated for each one, so no production data is required to evaluate them. After delivery, you self-host: there is no Nexgile cloud in the loop.
- Previews run on generated sample data, and Studio holds no credentials for your live systems
- You self-host the delivered software on your own infrastructure
- Nx IDE sends no telemetry, and works with local models so code can stay on your own hardware
- Data residency across our Herndon, VA and Hyderabad, India operations
- GDPR and CCPA aligned; retention and deletion on request
- Sub-processor transparency available on request
AI you can explain — and defend.
For an agentic-AI company, governance is a first-class concern, not fine print. Our foundation is built to make automated decisions accountable.
- Explainable, repeatable decisions — not “the model decided”
- Rule steps for routine work — the same result every run, with no AI call
- Human-approval gates assigned by role, team or variable
- An optional tamper-evident audit log in exported workers — specification available to your auditors
- Prompt-injection screening of business descriptions, plus optional guardrails and personal-data redaction in exported workers
Four principles, built to be checked.
Every claim on this site stands on one of these. Each is something you can check — in a live session, in an artifact your auditors request, or in a preview run — not something you are asked to take on faith.
Tested on sample data
Every worker can be run first on sample data generated for it — or files you choose to import — with a real AI model and a live, step-by-step trace. Approval steps pause the run until someone decides; evaluation suites, A/B tests and regression checks score every change. Studio holds no credentials for your live systems, so your security team can approve a description-based evaluation without a data-sharing agreement.
See it in Nx·StudioRules before AI
A rule step can’t be flipped by “are you sure?” or a prompt trick: it applies a fixed business rule, written in plain English, with no AI call — the same answer every run. AI steps are kept for genuine judgment; automatic gates check each case for policy, confidence, risk, cost or deadline; and human-approval steps hand decisions to a person.
Watch a worker sort decisionsReadable workflows
Every digital worker is a readable, versioned workflow, not a tangle of prompts: what it does, who approves what, where it escalates. Compare versions when policy changes, test them in preview, and hand the workflow to your auditors — it ships inside the export, so they read what the worker will do before it does it.
Governance you can readYours to run
Every export is a complete Python project with its runtime included, a STRIDE threat model, and an AI bill of materials — a machine-readable inventory of what’s inside, with a signed attestation your engineers verify offline, on your hardware, without trusting us. The code carries no licence keys and makes no calls to Nexgile, and a signed hash confirms the package you deploy is the one you reviewed. “No lock-in” stops being a slogan and becomes a property you check yourself.
Verify it yourselfDon’t take our word for any of this.
Every export ships an AI bill of materials — a machine-readable inventory of what’s inside — with a signed attestation your engineers verify offline, on your hardware, without trusting us. Sample export files are available on request: the verdict is yours to read.
And because each export carries a signed hash of its contents, you can confirm the package that is deployed is the one that was reviewed.
Determinism gets the same treatment — demonstrated, not asserted. In any live session, ask us to rerun the same scenario and watch every rule step land on the same answer.
- checkBill of materials the components and AI models it uses, listed
- checkSigned attestation verifiable offline, on your own hardware
- checkNo calls home no licence keys, no calls to Nexgile
Checked on your hardware. The verdict is yours to read — and the signed hash confirms the package you deploy is the one you checked.
Your auditors read what a worker will do before it does it.
Governance here isn’t a dashboard over a black box. It’s a set of artifacts you can read, diff, and keep.
The workflow is the audit artifact
Every worker is a readable, versioned workflow — its agents, tools, approval gates, routing and decision rules — not buried in prompts and not trapped in a vendor’s runtime. One artifact serves three audiences at once: your engineers version and compare it like code, your governance team reads it before deployment, your auditors check against it after.
An audit trail in the worker
Exported workers can keep a tamper-evident audit log — each entry chained and sealed — recording what the worker did and who approved what. Rule steps return the same answer every run, and cannot be talked out of policy by a clever prompt.
Approvals are designed in
Approval gates are part of the workflow — assigned by role, team or variable — and rehearsed in the preview. A reviewer-inbox app (still in progress) can ship with the export as code you own: your approval policy lives in the software, not in a setting on our servers.
How we handle the data you choose to share.
Previews run on sample data generated for each worker — or files you choose to import — and Studio holds no credentials for your live systems. It doesn’t need your production data to design, test or demonstrate a worker.
Design-time inputs are different, and we treat them that way: if you choose to share documents, process notes, or event logs to ground your workers in your real volumes, you decide what to share, it’s used only to design your workers, and a data-processing agreement is in place before anything moves.
A description-based evaluation shares nothing at all.
What’s deterministic, what isn’t — and who decides.
Live AI judgment is probabilistic by nature.
That’s why your workers use as little of it as possible: every decision that follows a fixed rule runs as a rule step that returns the same answer every run — and you can rerun your saved scenarios to check. Same input, same path, same answer.
Where judgment is genuinely needed, low-confidence cases can be routed to a human-approval step instead of letting the worker proceed on its own.
Escalation isn’t a courtesy. It’s a control.
What your auditors can request.
Every artifact below is sanitized and shared on request — write to security@nexgile.com.
A sanitized workflow excerpt
Rendered as a readable document — your governance team reads what a worker will do before it does it.
A sample evaluation report
Generated from a preview run on sample data — the evidence procurement will ask for.
A sanitized export in miniature
The project tree, its README and its tests from a sample export — what you receive, in practice.
A sample threat model
The STRIDE threat model every export includes — redacted to properties, never configurations.
The audit-trail specification
The events the optional tamper-evident audit log records and the properties it guarantees.
A sample AI bill of materials
With its signed attestation and the steps to verify it offline — the verdict is yours to read.
Your auditors can read what the worker will do before it does it.
Scoped on purpose. Kept on purpose.
One plain sentence each — scoped to what the platform proves, because a scoped pledge beats a falsifiable absolute.
Previews run on sample data — Studio holds no credentials for your live systems.
No vendor runtime. No per-decision billing. No calls home.
Nothing stops working if you stop paying us.
Every number on this site comes with how we measured it. When we can’t measure something honestly yet, we don’t publish it.
The house rule covers other people’s numbers too: third-party statistics appear only quoted, attributed, and dated. Our delivery counts carry a method footnote and are re-verified before every publication. And figures that stop reproducing are retired sitewide — not quietly kept.
Reliability
Because you self-host the software we deliver, its availability is in your hands — on your infrastructure, under your SLAs. Our own platform is continuously monitored, with a defined incident-response process. A public status page for Nexgile services is on our roadmap.
Talk to the right team
- Security & disclosure security@nexgile.com
- Privacy & data rights privacy@nexgile.com
- Accessibility accessibility@nexgile.com
You generate it. You preview it.
You own it. You run it.
Bring your security, risk, and compliance teams. Every evaluation runs on sample data, never on your live systems — so your security team can say yes sooner.