Opportunity portfolio
A ranked view of candidate workflows by economic value, frequency, process stability, data readiness, change burden, and consequence.
AI Audit
Find the first AI workflow worth funding.
We compare AI opportunities by their effect on revenue, margin, capacity, quality and risk. We examine the current workflow, establish a baseline and turn the strongest opportunity into a practical 90-day roadmap.
What you receive
Decide which AI workflow to fund.
First-workflow business case
The unit of work, baseline, value mechanism, implementation boundary, quality floor, risks, and assumptions for the recommended first move.
90-day roadmap
Sequenced discovery, prototype, controlled pilot, adoption, measurement, decision gates, owners, systems, and stop rules.
Authority and control design
What code may do, what AI may propose or execute, where people review or approve, and how failures are detected and recovered.
Data and system map
The systems, artifacts, permissions, integrations, and evidence needed for the first implementation.
Executive decision brief
Fund, validate further, fix the process first, or decline, with the reasoning and evidence needed to support the decision.
Inputs and participants
Assess the work as it happens.
| Needed | What it contributes | Typical participant |
|---|---|---|
| Leadership priorities | Revenue, margin, capacity, service, quality, hiring, or risk pressure | CEO, COO, CFO, functional executive |
| Workflow truth | Normal path, exceptions, workarounds, handoffs, judgment, and failure modes | Workflow owner and frontline operators |
| Systems and artifacts | Applications, records, documents, messages, policies, and permissions | Operations and technical owner |
| Current performance | Volume, touch time, wait time, rework, cost, quality, backlog, and outcomes | Metric owner or analyst |
| Decision authority | Who can approve scope, access, pilot controls, and adoption changes | Executive sponsor |
We do not require perfect data before the first conversation. Missing evidence becomes an explicit readiness gap; it does not become an invented number.
Process
Choose a first project with a clear business case.
1. Frame the outcome
Name the business pressure, decision owner, review window, and what evidence would make the audit useful.
2. Observe the work
Map the workflow, systems, actors, artifacts, exceptions, failure cost, and current performance.
3. Compare opportunities
Score value, volume, feasibility, data, change burden, risk, and measurement readiness across the shortlist.
4. Establish the baseline
Freeze the unit, denominator, formulas, sample window, owner, quality floor, and before/after comparison.
5. Design the controlled path
Define the smallest implementation, authority boundaries, approvals, evaluation, rollout, and recovery.
6. Make the decision
Deliver the roadmap and executive brief, then agree separately whether and how implementation proceeds.
Fit
Decide whether the audit fits your current needs.
Good fit
A real operating constraint matters now; leadership can name an owner; operators can participate; and the company is willing to measure the current workflow and control a pilot.
Not ready yet
No accountable owner, no access to the operation, a constantly changing process, or a request to deploy autonomously without evaluation and human control.
Not included by default
A generic AI workshop, company-wide transformation, software licenses, a production implementation, managed operation, or invented financial assurance. Those require separate explicit scope.
Before work begins, we agree the exact scope, participants, access, timing, fee, confidentiality, and final deliverables. The fee and any subsequent implementation depend on that agreed scope.
Questions
What buyers usually ask
Do we need to know which AI project we want?
No. The audit is designed to compare plausible workflows and identify which one has the strongest business case and readiness.
Is the audit a generic AI strategy presentation?
No. It is grounded in one or more real workflows, their operators, systems, artifacts, baseline measures, exceptions, and business outcomes.
Does the audit include implementation?
The audit produces the decision, baseline, control design, and roadmap. Any implementation scope is a separate explicit agreement.
How do security checks fit into the AI Audit?
The audit maps the workflow, data access, action boundaries, owners and evidence needed for an implementation decision. Agree any specialist security testing, remediation and retest as explicit deliverables with the relevant reviewers. See the scope and cost drivers for a launch review.
What is agreed before the audit begins?
We agree the scope, participants, access, timing, fee, confidentiality and deliverables with you before work begins.
First conversation
Start with the number leadership most needs to improve.
We will clarify the business pressure, candidate workflow, owner, and available evidence. If an audit is not the right next step, we will say so.
What are you trying to do?
See how the diagnosis shapes a build
The patient-access case describes a custom operating report assembled before the voice/SMS rollout. It distinguished calls answered, requests captured and actions completed. That is the kind of distinction an audit needs before selecting software.
Prepare with the prioritization guide and baseline guide. The exact deliverables and decision window are agreed for your engagement.