For businesses
AI and Automation Audit
Everyone's telling you to automate. Nobody's told you what to actually automate, or whether it's worth it.
You've been sold AI. Your team is experimenting with ChatGPT. You're paying for tools nobody uses. A vendor has pitched you on an agent that will "transform your operations." You don't know which, if any, of this is real.
This audit maps your actual workflows, identifies where automation delivers ROI, and gives you a prioritised plan. No tool pitch. No hype.
Why this exists
Why this service exists
Not enterprise AI projects. Not chatbots on your homepage. The real mid-market automation wins are the workflows that currently involve:
- Manual data re-entry between systems
- Report generation and formatting
- Client communications with predictable structure
- Approval chains and hand-offs
- Content operations (briefs, drafts, publishing)
- Billing, invoicing, and reconciliation
These are the workflows where modern AI and automation deliver real hours back, with reasonable implementation cost and low ongoing maintenance.
What I bring
Receipts and credentials
Most providers selling AI and automation jump straight to "let's build something." That works for them. It rarely works for the client. You end up with a bespoke build solving a workflow that wasn't the real bottleneck, or a tool that looked impressive in the demo and gathers dust in month three.
I map the opportunity before anyone builds anything. What your team actually does, where the manual work is, where the data lives, which workflows are worth automating, and which are better left alone. The right diagnosis sits at the intersection of business operations, data architecture, and technical feasibility. Then you get a roadmap you can act on, with or without me.
Two-stage structure
What this looks like in practice
At a large residential homebuilder I built the automation pipelines that cut 80% of the manual production work across the marketing team:
- Floorplan SVG processing. 30 minutes per floorplan, reduced to seconds.
- Image processing pipeline. Automatic whitespace detection, auto-crop, format conversion. Eliminated manual image prep entirely.
- API-driven publishing integrations. House and land package publishing: 30 minutes per package, handled automatically. Home design publishing: 2 hours per design, automated.
Running my own practice, I've built:
- Human-in-the-loop billing automation. Queries ClickUp for billable hours, Slack confirmation step for human review, dispatches to bookkeeping software, sends the client invoice. Multi-system orchestration, zero manual intervention after human sign-off.
- Production LLM integration. Connected a large language model to a headless CMS via a custom MCP server using the CMS GraphQL API. Live content operations driven by the AI against real CMS data.
Around two years of hands-on production AI use across OpenAI, Anthropic, and open-source models. Not prototyped. Built, deployed, used.
Tools and platforms
Frequently asked questions
-
What is an AI and automation audit?
A structured review of your actual workflows to find what's worth automating, what AI can genuinely do for you right now, and what it can't. You get a mapped list of opportunities with effort ratings, tooling recommendations, and a 90-day implementation roadmap. Audit first, tools second.
-
Why start with an audit instead of just trying AI tools?
Because the expensive failure mode is automating the wrong thing. Tools are cheap; rework and abandoned rollouts aren't. Mapping workflows first means the business case exists before any subscription does.
-
Which tools do you recommend?
Whichever fit the workflow, the budget, and your team's tolerance for maintenance: that genuinely varies. I'm platform-agnostic across the n8n, Zapier, and Make class of tools (here's the honest comparison) and conservative about bolting LLMs onto processes that needed a process fix, not a model.
-
What can AI actually automate in a business like ours?
Reliably today: document handling, data entry between systems, first-draft content, triage and routing, and reporting assembly. Unreliably: anything requiring judgement you'd not delegate to a new hire on their first week. The audit draws that line for your specific workflows, and this article covers the general pattern.
-
How long does it take?
Typically two to three weeks. Unlike the website and analytics audits, this one starts with discovery: I need to sit inside your actual processes and understand how work moves through the business before anyone can honestly say what AI or automation would help. Because scope depends on what discovery finds, pricing is confirmed after that first conversation rather than published as a flat rate.
Pricing
Pricing
From $2,500 for the standard audit: workflow mapping, opportunity prioritisation, written report, and a recorded walkthrough.
For enterprises with more complex workflows or multiple business units, custom scopes available on request. Discovery call is free.
Next
What happens next
The discovery session is the right starting point. Two hours, $750, applied against Stage 2 if you proceed.