Engineering,
verified.
Requirements, models, verification, schedule and risk on one traceable thread, with an AI co-engineer that drafts, audits and traces the work, then leaves the signature to you.
Free for 7 days, with welcome credits for every AI feature. No card required.
One thread from intent to evidence.
Every artifact carries a code and its links. Change a requirement and whatever depends on it is flagged as suspect until someone reviews it, while signed baselines keep the version you agreed to.
The whole V, in one workspace.
Ngenaire follows the systems engineering lifecycle from concept of operations to verification report. Each stage feeds the next, and the co-engineer works across all of them.
Requirements that read like engineering, not prose.
Capture the concept of operations and the statement of work, then write requirements in EARS. Quality checks flag ambiguity as you type, the co-engineer critiques and proposes rewrites, and baselines freeze what was agreed.
- EARS patterns with rule-based and AI quality scoring
- Outline, custom attributes and saved views
- Baselines and staged change requests with signed approvals
- CSV and ReqIF round-trip, DOCX and SRS export
A co-engineer that shows its work.
It reads the project, calls the same tools you use and leaves a trail. Every tool call is visible, every change lands in history, and nothing is approved without a person.
$ claude mcp add --transport http ngenaire \ https://ngenaire.com/mcp \ --header "Authorization: Bearer ngn_pat_…"
Works through tools
It reads and edits the project through the same services as the interface, under your role. Every tool call is shown with its arguments.
Specialists on call
Requirements analyst, V&V engineer, risk analyst, project controls and architect, each a slash command away.
Your client, your projects
Connect Claude Code, Claude Desktop or Cursor over MCP with a scoped, expiring personal access token.
Credits only when it works
AI actions are metered to whoever runs them. Editing, tracing and planning never spend credits.
Trust, with receipts.
Approvals are signed, actions are logged, tenants are isolated and tokens expire. The security page explains every control and how it is built.
Ngenaire holds no compliance certifications today and is not suitable for ITAR or EAR controlled data or classified information.
- Passkeys and recovery codes
- A WebAuthn passkey as the second step at sign-in, ten single-use recovery codes, and a lockout after repeated failed passwords.
- WebAuthn, argon2id
- Server-side sessions
- Opaque tokens in HttpOnly cookies with only a hash stored. Idle and absolute expiry, and every device can be signed out.
- HttpOnly, SameSite=Lax
- Signed approvals and baselines
- Approving a requirement, a change-request stage or a baseline records who, when and in what role, with a signature over the content.
- signed
- Least-privilege roles
- Owner, editor, approver and viewer on every project. Strict approvals take sign-off away from editors entirely.
- strict approvals
- Tenant isolation
- Every query is scoped to its project. Another tenant’s object is indistinguishable from one that does not exist.
- 404, never 403
- Audit trail
- Sign-ins, second-factor changes, role changes, approvals and access tokens are recorded in the governance log and your security activity.
- governance log
- Scoped, expiring tokens
- Personal access tokens for scripts and MCP clients are hashed, read- or write-scoped, and expire in 7 to 365 days.
- ngn_pat_, read or write
Every plan is the whole product.
Plans differ only in how many AI credits they include. Writing, tracing, baselining and scheduling never spend credits.
The trial never converts on its own: you choose a plan when it ends.
What engineers ask first.
Something else on your mind? An engineer reads every message to our solutions inbox, or browse the documentation.
Who is Ngenaire for?
Systems engineers and the programs around them: aerospace, defence, automotive, robotics, energy and any hardware-heavy effort that lives on requirements, verification evidence and a schedule someone has to defend. It speaks EARS, RVTM, ConOps, TADI, baselines and critical path natively.
Does the co-engineer change things without asking?
It works through the same tools the interface uses, under your role and permissions, so it can only do what you could do yourself. Approvals and baselines are always yours: a signature records who approved what, in which role, and when.
Is our project data used to train AI models?
No. Project content goes to our AI provider only when you use an AI feature, and only what that request needs. We never use it to train models.
Can I use it from Claude Code, Claude Desktop or Cursor?
Yes. Create a personal access token with read or write scope and an expiry, then add the Ngenaire MCP server to your client. The client works as you, with exactly your permissions.
Can we bring our existing requirements?
Requirements round-trip through CSV and ReqIF, the interchange standard other requirements tools speak, and export to DOCX or a generated SRS document. Nothing is locked in.
Is it suitable for export-controlled or classified data?
No. Ngenaire is not certified for ITAR or EAR controlled technical data, or for classified information. Keep that material out of your workspace.
Bring your next program into one workspace.
Start with the sample autonomous-vehicle program, or point the co-engineer at your messiest requirement set on day one.
Free for 7 days, with welcome credits for every AI feature. No card required.