AI audits
We interview the people doing the work, assess how the business operates today, and turn what we learn into scores, a blueprint, and a prioritized action plan.
How an audit works
The process is interview-based. For a small team, we audit the core group. For a larger company, we go department by department.
Meet with leadership to understand the business, team structure, current tools, and major operational pain points.
Talk with the people doing the work and document the steps, time, tools, handoffs, and friction.
Assess the findings, rank the opportunities, and deliver the scores, blueprint, and action plan.
The AI efficiency audit
We fit the KPAI framework to your business. Governance wraps all five layers so access, ownership, review, and accountability are built into the plan.
01
How your team reaches and works with AI.
02
The models that power the work and can change as the market does.
03
The knowledge, SOPs, and context that make AI useful to your company.
04
The people or systems responsible for carrying the work through.
05
The business systems the stack reads from and acts through.
The automation audit
Each opportunity is scored and ranked, then tagged by who can build it and how much it lifts your automation potential. A few examples.
Builds a monthly list of clients hitting a home-purchase anniversary and drafts the check-in to send.
Monitors due dates and sends escalating reminders before things slip.
Turns a completed intake into structured briefs for the team.
What you receive
A representative view of the deliverable. The full blueprint and prioritized action plan are shared privately with each client.
Scored across the five layers, with governance around all of them.
FoundThe team uses one chat assistant inside a closed ecosystem.
RecommendAdd a CLI for heavy work and agent mode for tasks that touch your tools.
FoundStrong models already sit inside the tools you use, with no approved list.
RecommendReserve top models for hard tasks and use cheaper ones for routine work.
FoundKnowledge lives in a few people's heads and scattered, outdated documents.
RecommendWrite SOPs by department so the common work is captured and reusable.
FoundPeople do nearly all the work; one AI-assisted step is already in place.
RecommendTurn those SOPs into skills that bridge human and agentic work.
FoundYour CRM, finance, and file systems have APIs, but almost nothing is connected.
RecommendReplace tools with no API, then connect the rest to one system of record.
FoundThere are no formal rules for what AI can access.
RecommendSet three data levels and a testing practice for new tools.
This is an abbreviated sample. The full Standard Audit is shared privately with each client.
Start with the right question
We will talk through the situation and recommend the right level of assessment.
Book an audit consultation