CA (CAANZ) · 20 years audit & operations
Your operations, untangled.
Your capacity, ×multiplied.
AI operations consultant for Australian businesses — process automation, workflow optimisation, and AI integration.
More time building the business. Less time running it.
You own everything · No lock-in
Free setup
No consulting fees to start.
Subscription only.
Subscription only.
1 week
Your AI executive assistant,
working in days.
working in days.
0%
Tried AI, only 14%
have fully integrated it
have fully integrated it
0%
Of work is coordination,
not value
not value
Built on Anthropic
Integrates with
XeroGmailGoogle CalendarGoogle DriveHubSpotOutlookSharePoint
XeroGmailGoogle CalendarGoogle DriveHubSpotOutlookSharePoint
The Problem
Your team spends more time on overhead than on the work itself.
Writing specs so someone who wasn't in the room can act on a decision. Sitting in meetings so eight people can synchronise. Preparing decks so an executive can make a call. Filing tickets so work can be tracked across people who can't see each other's progress.
None of this is the value. The shipped product is the value. Everything else is the cost of producing it with humans who can't share context directly.
0%
of employee time spent communicating, not creating
Microsoft Work Trend Index 2023
0%
of work is "work about work," not value creation
Asana Anatomy of Work Index 2021
3x
increase in meetings per week since 2020
Microsoft Work Trend Index 2023
Microsoft Work Trend Index 2023 | Asana Anatomy of Work Index 2021
You're Not Alone
Most businesses are in the same place you are.
0%
of small businesses cite lack of technical expertise as the barrier
Goldman Sachs 10,000 Small Businesses, 2026
0%
have tried AI, but only 14% have fully integrated it
Goldman Sachs 10,000 Small Businesses, 2026
0%
say lack of expertise is the top barrier. Not cost — complexity.
Bookipi AI Adoption Report, 2026
You've probably tried ChatGPT for emails or marketing. That's the easy part. The hard part (the operations, the finance, the coordination) is where I work.
Goldman Sachs 10,000 Small Businesses 2026 | Bookipi AI Adoption Report 2026
The Framework
Understand, Optimise, Scale.
Three tiers of ambition for what I can do with your business. Each one earns the next.
Tier 1
Understand
"I can't fix what I can't see."
Before anything gets built, I see how work actually moves through your business. Not what you think happens, but what actually happens. The gap between those two versions is where time and money disappear.
Tier 2
Optimise
"Make what you have work properly."
Eliminate what shouldn't exist. Restructure what's broken. Then, and only then, automate what survives. In that order, always. I won't sell you AI where you don't need it.
Tier 3
Scale
"Do things you couldn't do before."
Not "how do we make this faster?" but "what would this look like if we designed it from scratch?" Your 30-person firm operating like 100. The work you've been turning away because you didn't have the capacity.
Most clients start at Understand and stay in Optimise for months. That's fine. There's real value in a well-optimised business. Scale enters when trust is established and you show ambition.
How I Optimise
I'll tell you to stop doing things before I sell you AI.
1. Eliminate
Does this process need to exist at all? Reports nobody reads. Approval chains for trivial amounts. Meetings that should be emails. Eliminating waste is free and immediate. This catches 20–30% of identified waste with zero technology.
2. Restructure
For processes that must exist: is the design right? Fix broken handoffs and sequences. Or restructure a role so AI handles the surrounding admin and the person's capacity multiplies. A PM handling 8 clients could handle 20.
3. Automate
For well-designed processes with high volume and digital data: build the skill, connect the tools, let the AI employee handle it. Not everything. The specific things where automation makes sense after the first two steps have cleared the ground.
Free up your people. Scale your business.
25% → 65%
High-value work
Client conversations, commercial decisions, strategic thinking. The work your best people were actually hired to do. After ERA, it's not squeezed into the margins. It's the majority of the week.
Target: +160% capacity
35% → 15%
Coordination overhead
Meetings that exist because no one has the full picture. Status updates, handoffs, "just checking in." Most of this disappears when the AI employee holds the shared context.
Target: -57% reduction
25% → 10%
Assembly work
Reports nobody enjoys assembling. Data pulled from three systems into a spreadsheet. The bank report that takes two days. The AI employee does the assembly. Your ops manager reviews and signs off.
Target: -60% reduction
15% → 10%
Process overhead
Approvals for trivial amounts. Chasing people for signatures. Rework from unclear instructions. Some of this gets eliminated entirely. The rest gets restructured so it takes minutes, not hours.
Target: -33% reduction
What Changes
What this looks like in practice.
Typical scenarios from ERA diagnostic modelling. Same people, same tools, different capacity.
Your Consultant
1 client pack/day
4 client packs/day
AI handles data extraction, formatting, and report assembly. Your consultant focuses on analysis and client conversations.
Your Project Manager
8 clients
20 clients
Status updates, scheduling, and progress reports are automated. The PM manages relationships, not spreadsheets. Clients feel more attended to, not less.
Your Ops Manager
2 days on the bank report
20 minutes
Data pulled from Xero, formatted to lender requirements, ready for review. The ops manager checks and signs off instead of assembling from scratch.
Ready to get started?
30-minute call. Tell me what's not working — I'll tell you if I can help.
Start With a Free AI Executive Assistant
What I Actually Say
I'll tell you things other consultancies won't.
Most of what you've heard about AI is from people selling AI. Here's what I actually say in discovery conversations.
"We're not ready for AI yet."
Good. That's honest. Most businesses aren't, and the ones who jump straight to automation usually automate the wrong things. I start by mapping how work actually flows. That's valuable whether or not you ever use AI.
"We tried AI tools and they didn't stick."
They probably solved a tool problem, not a process problem. If the underlying workflow is broken, better software just produces broken results faster. I fix the process first. The AI is the last step, not the first.
"What about our data? Where does it go?"
Your files, credentials, and tool accounts stay on your machine. Conversation data is processed by the AI provider's servers — the same trust model as any cloud tool like Xero or Gmail. Their commercial terms prohibit training on your content, and all data is encrypted in transit and at rest. I don't hold your data and I don't share it between clients. For regulated industries, I can route through your own AWS or Google Cloud.
"We don't want to be locked into a vendor."
You won't be. You own the deployment. If you stop paying me, your AI employee keeps working. The retainer is for ongoing improvement: new skills, edge case handling, the thinking about what to build next. Not for keeping the lights on.
"AI will replace our team."
I free up your best people so they can do more of what they're good at. A PM handling 8 clients could handle 20. Same person, different capacity.
"How do we know this will actually work?"
You don't. Anyone who guarantees it is selling you something. What I guarantee is that by the end of the discovery call, you'll see your operations more clearly than you did before. That clarity is valuable on its own. Everything else builds from there.
How This Is Different
I'm not an AI agency.
If you've been looking at AI consultancies, you've seen the pattern: a discovery phase, a roadmap, a build sprint, and then they leave. Here's how this works differently.
| Typical AI agency | AI Workflow Advantage | |
|---|---|---|
| Delivery | Sprint: build, hand over, move on | Continuous retainer: build, iterate, improve every week |
| First value | Weeks to months of discovery first | Within a week. Your assistant has visibility across your team from setup day. |
| Coverage | "Here's a spec, we'll build to 100%" | 80% in the first week. 90%+ through real use. Edge cases surface naturally. |
| Framework | "AI readiness assessment" | Understand, Optimise, Scale. ERA decision filter inside Optimise. |
| Honesty | "You need AI everywhere" | "You don't need AI here. Stop doing this entirely." |
| Credibility | Tech background | CA + 15+ years audit, finance, operations. Speaks the owner's language. |
| Ownership | Vendor lock-in, proprietary platform | You own the deployment. Stop paying, it keeps working. |
| What you get | A dashboard nobody opens | An AI employee that does the work. Finished artifacts, not features. |
How It Works
Your AI employee is working within a week. Then I map what to build next.
We Have a Conversation
30 minutes · Free
You tell me what's costing you time. I map it, challenge it, and tell you honestly what's worth automating, and what you should just stop doing. No pitch. If it's not a fit, I'll say so.
Your AI Employee Is Working by Friday
Free setup · ~1 week
I set up an AI assistant on your computer and connect it to your email and calendar. Morning briefings, inbox triage, meeting prep, email drafting, meeting follow-up. You pay the platform subscription ($20–$100/month). I configure everything at no charge. No commitment. Keep it regardless.
I Map Your Business
Operations diagnostic · $5,000–$7,000
Once the AI employee has earned your trust, I go deeper. AI-conducted interviews with your team, cross-person triangulation, and every process through ERA: Eliminate what doesn't need doing. Restructure what's broken. Only then automate. The diagnostic tells me exactly what to build and in what order.
I Build What Your Business Needs
Retainer · $1,000–$3,000/month
From the diagnostic findings, I build capabilities that grow in complexity: quick wins first (invoice chasing, quoting), then process systems (client onboarding, reporting), then full business orchestrations (legal review, CRM management). It keeps getting better because someone who understands your business is actively improving it.
The AI Employee
It does the work. You review the output.
It sits on your office computer. Connected to your tools. You talk to it in plain English. It acts.
- "Chase the overdue invoices." It queries Xero, composes reminders in your tone, sends via Gmail, schedules follow-ups, and reports back.
- "Generate a quote for the Smith job." It pulls measurements, looks up rates, assembles the document, and drops it in your review queue.
- "What's my cash flow looking like?" Receivables, payables, recurring expenses, all pulled from Xero. Anything that needs attention gets flagged.
- "Prepare everything for the board meeting." Reports gathered, pack assembled, missing items flagged. Ready by morning.
What it connects to
Xero — invoicing, AR, cash flow, compliance
Gmail / Outlook — email composition, drafts, sending
Google Calendar — scheduling, reminders, briefings
Google Drive / SharePoint — file management, reports
CRM — lead follow-up, pipeline, customer data
You own the deployment. If you stop the retainer, it keeps working. The retainer is value, not lock-in.
What It Actually Does
Real problems. Built from how you actually work.
Every business has its own version of these. The discovery call is where you tell me which ones are costing you the most.
Invoices sitting unpaid
Invoices go out and then nothing. $80K outstanding and nobody's chasing because everyone assumes someone else is doing it.
The AI employee checks Xero every morning. Day 1: friendly reminder. Day 7: firmer follow-up. Day 14: final notice draft for your review. Day 30+: flags it for you with options.
Quotes take days
A prospect asks for a quote on Monday. By Friday you've lost the job because three competitors responded the same afternoon.
You say "quote the Smith job." It pulls measurements, looks up rates, assembles the document, and drops it in your review queue. Minutes, not days.
Leads go cold
New leads come in from the website or referrals. Nobody follows up within the first day. By the time someone calls back, they've hired a competitor.
Daily at 8am, the AI employee checks for new leads with no follow-up. Drafts a personalised email referencing their enquiry. Runs a Day 2, 5, 10 sequence until they reply.
No idea what's happening today
You walk in at 8am and spend the first hour checking emails, calendar, Xero, and Slack to figure out what needs attention.
Morning briefing lands in your inbox before you arrive. Cash position, overdue invoices, today's meetings with context, urgent emails flagged, and what the AI employee handled overnight.
Staff typing data from PDFs
Someone spends hours every week retyping information from supplier invoices, delivery dockets, or compliance certificates into spreadsheets.
Drop the PDF. The AI employee reads it, extracts the fields, and populates your spreadsheet or system. No templates to configure. It reads like a person would.
Board meetings are a scramble
Two days before the board meeting, the ops manager starts assembling numbers from five different systems into a pack that should have been done last week.
Monthly scheduled task. The AI employee pulls financials from Xero, project status from your tracker, and KPIs from your data. Assembles the board pack. Your ops manager reviews and refines.
Everyone asks Marcus
One person holds all the institutional knowledge. When they're on leave, everything slows down. When they quit, years of context walks out the door.
The AI employee becomes the knowledge base. It reads your documents, processes, and history. Anyone can ask "how do we handle returns from interstate?" and get the answer Marcus would give.
Best candidates gone before you finish
You post a role, get 60 applications, and take two weeks to review them. The best three candidates accepted other offers five days ago.
Drop the resumes. The AI employee reads all 60, ranks against your criteria, and gives you a shortlist with reasoning within the hour. You interview this week, not next month.
Cash flow is a mystery
You check the bank balance but have no idea what's coming in or going out. BAS due dates sneak up. Cash runs tight and you don't see it coming.
Ask "what's my cash flow?" It pulls receivables, payables, and recurring expenses from Xero. Projects 12 weeks out. Flags the week payroll and BAS collide.
Monday meetings waste everyone's morning
Two hours every Monday catching each other up on last week. Half the room is waiting for their turn to talk. The other half is checking email.
Before the meeting, the AI employee compiles what happened last week: completed tasks, open items, blockers, key numbers. Everyone reads a 2-minute summary. The meeting becomes 30 minutes of decisions, not 2 hours of updates.
Leo Malan
CA (qualified 2009) · AI Workflow Advantage
Not an AI agency. An operations improvement consultancy that uses AI as one of the tools. The credibility isn't tech. It's 15+ years understanding how businesses actually work, where money leaks, and what to fix first.
- 15+ years audit, corporate finance, operations
Chartered Accountant (CAANZ)- Perth based, AU businesses
- 15–200 employee companies
Common Questions
Straight answers.
30 minutes. You talk about what's frustrating you — the bottlenecks, the overhead, the things that should be simpler. I listen, ask questions, and tell you honestly whether I can help. You walk away knowing if this is worth pursuing. No pitch deck, no pressure.
Your executive assistant is running from setup day. I spend 15 minutes understanding how your team coordinates — where tasks live, how work gets tracked — and configure the assistant around that. Morning briefings, inbox triage, meeting prep start within the first week. Over the first month, I refine based on real usage. By month two, you have full visibility across your team and the assistant is surfacing what needs your attention before you have to ask.
Your AI executive assistant is free. I set it up, connect your email and calendar, and you're using it within a week. You pay the AI platform subscription ($20–$100/month) directly. No consultant fees, no commitment. Once the assistant has earned your trust, the operations diagnostic ($5,000–$7,000) maps how work actually moves through your business — what to eliminate, what to restructure, and what to automate. The ongoing retainer ($1,000–$3,000/month) then builds capabilities from those findings: simple automations first, then process systems, then full business orchestrations. A team recovering even 10 hours per week at $50/hour saves $26,000 a year in capacity.
Xero, Gmail, Google Workspace, Google Calendar, Google Drive, SharePoint, Outlook, HubSpot, and most CRM systems. The minimum useful stack is Xero + email + calendar. During the discovery call I map your specific tools to determine exactly what the AI employee can connect to.
Your files, credentials, and tool accounts (Xero, Gmail, etc.) stay on your machine. Conversation data is processed by the AI provider's servers — the same way Xero processes your financial data or Gmail processes your email. No shared database between clients. Each deployment is fully isolated. The AI provider's terms prohibit training on your content, with AES-256 encryption at rest and TLS in transit. For regulated industries, I can route through your own AWS or Google Cloud infrastructure with compliance controls you manage.
Your AI employee keeps working. Everything is on your machine, your account, your credentials. What you lose is the consultant's thinking: new skills, process improvements, and the ongoing conversation about what to eliminate, restructure, or automate next. Some clients graduate. That's the success story, not churn.
Best fit: Australian businesses with 15-200 employees, $5-50M revenue, where the owner knows things could run better but can't see exactly where. Not a fit: businesses below 15 employees (not enough complexity), businesses in financial distress (fix that first), or businesses that want validation rather than honest feedback.
Every external action starts on "confirm." The AI employee describes what it wants to do and waits for your approval before acting. I only move actions to automatic after you've seen them work reliably over weeks. Financial actions above a threshold you set are always blocked. There's also a hard safety layer the AI cannot override. It can't delete files, bypass permissions, or change its own rules.
Yes, and no. It's a chat window on a computer. Your team types what they need in plain English. "What's the status on the Henderson job?" or "Generate a quote for 200m of fencing at the new rate." If they can send a text message, they can use this. No dashboard to learn, no manual to read.
Your consultant reconnects the AI employee to the new tool and updates the affected skills. This is a normal part of the retainer. Businesses change, and the AI employee adapts. The business logic and skill instructions usually carry over; only the tool connection changes.
Software gives you a fixed set of features. A developer builds exactly what you ask for, then moves on. This is an operations consultant who uses AI as one of the tools. The value isn't just the technology. It's knowing what to automate, what to eliminate, and what to fix by changing the process instead of adding more tech. Sometimes the answer is to stop doing something entirely. That's not something a software vendor will tell you.