Start with the smallest engagement that can produce a defensible decision.
Qualify one workflow first. Expand discovery only when the customer needs it. Implement only when the workflow, observations, recovery, authority, and economics support proceeding.
Workflow Qualification
For a buyer who has one plausible workflow and needs a bounded build, narrow, defer, or no-go decision before funding broader discovery.
Trigger, authority, real operator steps, dependencies, exceptions, and likely blast radius.
One intended outcome, protected invariants, credible observations, and stop conditions.
Surviving recovery path, required identities, customer participation, and prerequisites.
Qualification record, implementation range, sample evidence design, and explicit no-go conditions.
Full Assurance Assessment
The assessment stands alone. It is not a sales document with a proposal attached. If the recommendation is “your process is sound, spend your money elsewhere,” that is what it will say.
A documented comparison of the agreed source fields and representative running state, with observed discrepancies and evidence limits recorded.
How a change actually travels from request to production today — including the steps that are informal, undocumented, or done by one person.
The specific ways a bad change reaches production undetected, ranked by likelihood and blast radius.
Hard gates, protected invariants, stop conditions, dependencies, and risk-reduction priorities.
Build, narrow, defer, or do not proceed — usable with or without OperiProof.
The Implementation
Taken only if the assessment justifies it, and only for one workflow at a time. We deliberately start with change validation, configuration compliance, or provisioning — never a platform rewrite.
One bounded customer workflow engineered and qualified end to end before production use
Validation that runs before anything touches the network, and an approval bound by digest to the exact configuration reviewed
Declared outcomes tested using agreed observations
A tested, workflow-specific recovery path with separate authorization
A reproducible workflow your team can operate and verify
Written for your engineers, not for us
A machine-readable record of changes executed through the qualified workflow, with a human-readable report generated from it and a SHA-256 manifest your team can re-check independently. Cryptographic signing is available where a record must be verified outside your systems. This is the artifact you retain for operations, leadership, or audit review.
Ongoing Assurance
Compatibility review, evidence and control-health review, recovery rehearsal, and bounded improvements may be contracted after an accepted workflow establishes real demand. This is not a 24/7 NOC or an unlimited incident service.
Qualification and full assessment use fixed launch prices when the boundary is known. Complex brownfield discovery uses time and materials with a written ceiling. Implementation is milestone-priced. Public prices remain off the site while the first quotes and engagements establish market evidence.
The fifteen-minute technical demo
A reproducible FRR 10.2.1 and containerlab reference workflow catches a deliberately unsafe static-route proposal before device access, executes a separately approved safe candidate, detects a post-change assurance failure, performs separately authorized recovery, and independently verifies the retained evidence. It uses synthetic data and is not customer production qualification.
The completed reference demonstrates the full assurance sequence. Customer platforms, identity, sources, observations and recovery are qualified during an engagement.
Request the fifteen-minute demoStated plainly, because the boundaries are the credibility.
No trenching, splicing, cabling or construction.
No 24/7 on-call, no NOC, no help desk.
Nothing changes production without a human approving the specific change.
We start with one workflow.
If the right answer is a tool you can run yourself, that is what we will say.