Service · lead engagement

AI Workflow Enablement: the four-day First Workflow.

Four days. One workflow shipped. A runbook you keep either way. Aligned to NIST AI RMF and ISO/IEC 42001. The same governance discipline Cyber Node brings to cyber risk, applied to your AI deployment.

First Workflow

The First Workflow engagement is fixed in scope, duration, and price. It ends at the end of week two unless we both agree there is more value to capture.

A$6,000 + GST

Four working days. First measurable improvement by week two. Payable in full on signature.

Book a scoping call

What you get

  • Process map across the function or business unit in scope.
  • AI-suitability triage of each mapped workflow.
  • One workflow improvement, deployed end-to-end, with explicit guardrails and an escalation path.
  • Runbook covering operation, escalation, and shutdown.
  • Measurement note: before-and-after metric agreed in scope.
  • 60-minute verbal debrief.

Day by day

1

Discovery

Process inventory across the function in scope. AI-suitability triage. Scope sign-off at end of day.

2

Deep-dive

Workflow selected. Agent specification. End-of-day gate to narrow scope if Day 3 build will not fit.

3

Build

One workflow built end-to-end with guardrails and escalation gates tested.

4

Ship

Deploy. Measure against the before-and-after metric agreed on Day 1. Runbook handover and verbal debrief.

Day 3 in detail

Day 3 is the day you are buying. Days 1, 2 and 4 set it up, measure it and hand it over. Day 3 is the day a workflow agent that did not exist on Monday morning ships into your environment with named guardrails and a working escalation path.

The build covers three things, every time. What gets built: the agent itself, configured against the workflow scoped on Day 2, with the connectors it needs (email, calendar, CRM, file system, accounting system, whatever the workflow requires), the prompts and tool definitions checked into version control, and the runbook section that documents how it operates and how to turn it off. The guardrails: scope boundaries enforced in the prompt and in code so the agent cannot drift into work it was not built for; rate limits and cost caps so a runaway loop cannot generate a four-figure cloud bill overnight; logging and tracing so every decision the agent makes is auditable. The eval: a small evaluation set of representative inputs with expected outputs, run against the agent before deployment and again after, with the results recorded so you can detect drift the next time you change the prompt or the model.

Day 3 closes with the agent ready to ship on Day 4. The eval set, the guardrails and the runbook are checked in alongside it. None of those three are optional. If the eval fails or a guardrail is missing, the workflow does not deploy and the engagement scope tightens on Day 4 to land something narrower that does meet the bar. We do not ship on hope.

What happens after

The engagement ends at end of week two unless we both agree there is more value to capture. You leave with the improvement and the runbook either way.

If the first improvement returns the investment, we can scope a second workflow. Engagements can grow into multi-workflow deployments or ongoing oversight, but those conversations happen after the first improvement ships, not before.

This service is not

  • An AI strategy deck.
  • Bespoke ML model development or training.
  • An open-ended retainer with no shipped artefact.

Frequently asked

First Workflow is the only fixed-price engagement we sell from a price list. Everything above it is scoped to your business after the First Workflow has shipped. The price is real, but it is set against the work, not the buyer.

Configurations are documented in the AI model register. The providers used during your engagement are named in the SOW and disclosed in the configuration disclosure standard delivered with the work. The marketing site does not name providers as a matter of voice policy.

No. Bespoke ML model development and training are out of scope. Core Nova works with deployed and pre-trained AI capabilities (LLMs, agent frameworks, AI features inside SaaS products) and helps clients integrate them safely.

No. The First Workflow deliverable hands the working improvement to your engineering or operations team. Ongoing operation sits with you. Ongoing oversight (drift review, policy maintenance) is a separate retainer engagement scoped after the first improvement ships.

Yes. AI incident response is in scope for engagements where it applies, and can be purchased standalone at A$2,000 per day. Same incident-response posture Cyber Node brings to cyber incidents.