Actuary · Builder · Melbourne

I find what's costing you time and money,
then build the fix.

A one-person practice for founders, operators and teams inside larger businesses. I inspect what you already run, find where it leaks, then build the fix straight into your existing stack. An actuary who ships production software. One accountable person, start to finish.

Start a conversation See the work
Senior and accountable. Concept to deployed in weeks, not quarters.
Gap scan / a typical first look Scanning
Where time and money are leaking Ranked by impact
Ad spend with no clear attributionBudget moving on a hunch across channels High
Pricing set once, never revisitedMargin left on the table on every order High
Hours a week on manual reportingA task software should be doing for you Automate
Reconciliation done on trustPayments leaking quietly at the edges Medium
Highest-impact fixes identified: 2 · build time: ~weeks Integrates with your stack
Every engagement starts here. Find the leak, size it in dollars, fix the biggest one first.
ShopifyMetaGoogleKlaviyoGA4StripeAzureAWSSQL Server
Built into the stack you already run. Nothing bolted on the side.
The problem

Most advice stops at the slide deck.

You pay for a recommendation, then go looking for someone to build it. Half the value dies in that handoff, and the other half waits in a backlog.

The deck isn't the fix. The system is the fix.

What I do

One person, three lanes.

Because the highest-impact fix usually crosses all three.

Build
AI & automation
Scoped through deep discovery, built to a clear framework with guardrails, wired into what you already run.
AI use-case discovery & scoping
Agents & workflow automation
Guardrails, evals & governance
Grow
The growth stack
Getting found, getting measured, and turning more of the traffic you already pay for into revenue.
SEO & local search
Email, Meta & channel strategy
Analytics & attribution
Solve
The hard numbers
The data problems that put a real figure on the table and de-risk the next decision.
Pricing strategy & modelling
Payroll & commission analysis
Forecasting & scenario analysis
Who you're hiring

An actuary who builds the software. Both, in one person.

I trained as an actuary, one of the most demanding quantitative qualifications there is, then spent eight years shipping production software.

The mind · actuarial

Trained to price risk and quantify the uncertain.

I turn messy, incomplete data into defensible numbers your finance team will accept, model what happens before you commit a dollar, and find the few decisions that move the result.

Probabilistic modelling Risk & pricing Statistical analysis Forecasting
The hands · engineering

And the ability to actually ship it.

Eight years of production systems, in the languages and cloud you already run. I don't hand over a recommendation and leave. I build the system that captures the value, with no junior team learning on your budget.

C# / ASP.NET Advanced SQL Server Advanced R AWS & Azure AI / LLM integration
No handoff. The person who finds it, sizes it and builds it is the same person.
The work

Three products. Blank file to production. Solo.

Not theory and not slideware. All three are live: scoutr.world, zyro.world and betbetter.world.

7stage nightly AI engine, built end to end
Scoutr · orchestrated-AI marketing · scoutr.world

An AI marketing team that never sleeps.

Every night it reads the market, reconstructs the buyer from real public language, tests campaigns against that buyer before any spend, ships them across six connected platforms, and compounds on the results. A seven-stage orchestrator with guardrails, evals and autopilot.

Built solo6 platform integrationsLLM orchestrationBrand-safety guardrails
The proof for "AI in production". Scheduled, fault-tolerant, spending real money safely.
Zyro · attribution & CRO · zyro.world

Tracking the revenue everyone else misses.

Server-side tracking that survives ad blockers, and intent-based tests that pick the right page for each visitor source, including the AI traffic most tools ignore. Concept to a working product with a production AI layer.

Built soloServer-side attributionProduction AI layer
AIattribution engine with a production AI layer
The proof for "measure what's real". Attribution that holds up when the pixel doesn't.
Edgea sports analytics platform, modelled from scratch
BetBetter · sports analytics · betbetter.world

Turning raw signal into decisions.

Scrape the real signal, model it honestly, automate the read. The same probabilistic instinct that runs through everything else, pointed at one of the hardest prediction problems there is.

Built soloPredictive modelsAutomated pipeline
The proof for "model it properly". Honest accuracy, automated end to end.
My clients

Real businesses. Real numbers.

Animal Emergency AustraliaBusiness intelligence across a ten-clinic emergency vet group: pricing methodology, payroll automation and a commission audit traced to the cent.Solve
Eyezone EmporiumFull analytics and merchant setup for a luxury eyewear boutique, then local SEO and marketing built on clean data.Grow
Vintage Posters OnlyChannel and marketplace strategy, plus email and Meta marketing to turn an audience into repeat buyers.Grow
Fragrance ArtisansSEO, growth and marketing for an Australian fragrance brand, built to be found by the right buyers.Grow
Stockroom CafeLocal search for a Pyrmont cafe. Homepage rebuild, a catering page, and a post-launch plan.Grow
PropkornBusiness growth strategy, working the levers that move revenue from first principles.Grow
What I'm capable of

Actuary-grade rigour, across the whole stack.

Actuaries were the original data scientists: the same statistical core that underpins modern machine learning, with the added discipline of making a model defensible to a board.

01
Data analytics & BI
Advanced SQL Server and advanced R. Right now, the BI seat for an emergency-vet group across ten clinics, turning operational data into pricing and payroll decisions.
SQL Server · R · live with a client today
02
Data science & machine learning
GLMs, gradient-boosted models, survival and probability models, built, validated and deployed — from pricing engines to a sports-analytics platform modelled from scratch.
Trained, monitored, held to honest accuracy
03
AI & LLM systems
Agent orchestration, evals, guardrails and brand-safety. A seven-stage nightly engine running in Scoutr and a production AI layer in Zyro, with real spend behind them.
Production, not prototypes
04
Pre-spend evals & guardrails
Campaigns tested against a reconstructed buyer before a dollar is committed. Controversies and competitor crises blocked from paid spend automatically. The judgment is in the system.
An eval harness for marketing
05
Financial analysis & forecasting
Pricing models, scenario analysis and forecasts built to survive scrutiny. A ten-clinic pricing rebuild and a commission audit that recovered real money are the recent examples.
Numbers your finance team will accept
06
Risk & uncertainty
Trained to price risk and quantify what's uncertain. I size the downside before you commit, and tell you which decisions actually move the result and which don't.
The actuarial core
How I work with you

Discover. Scope. Build with guardrails. Govern.

The biggest predictor of return is choosing the right use case. So we start there, put a number on it, and never ship without guardrails and an owner.

01 · Discover

Find the leak

I map your stack, data and workflows to surface where AI and automation genuinely pay off, and where they don't.

02 · Scope

Put a number on it

Each use case gets a defensible figure, a clear scope and success criteria, ranked by impact. We start where the risk is lowest.

03 · Build

Ship it safely

Built with the controls baked in: human-in-the-loop, evals, fallbacks, audit trails. Integrated and proven against a baseline.

04 · Govern

Hand it over

Monitoring, documentation and clear ownership, so it keeps running safely long after I'm gone.

Senior and accountable NDA-friendly Works within your security & data rules No technical debt
Fair questions

The things you are right to ask.

Why one person instead of an agency?

Because the handoff is where value dies. With an agency, the person who scopes the work, the person who builds it and the person accountable for it are three different people. Here they are the same person, so nothing is lost in translation and there is one name on the result.

Will it work inside our security and data rules?

Yes. Everything is built to slot into the stack you already run and to work within your security and data constraints, with guardrails, audit trails and clear ownership built in. No new platform to adopt, no technical debt left behind.

What does it cost?

Each build is scoped and priced against the leak it closes, so the number on the proposal is smaller than the number it recovers. The first conversation and the first look at your setup are free, and I will tell you straight if there is nothing worth building.

Can't we just use ChatGPT or Copilot?

Use them, they are great drafting tools. But a chat window can't choose the right use case, wire itself into your systems, run unattended with guardrails, or be accountable when something breaks. That last mile is the work, and it's the part I do.

If something should be working better, let's find it.

Tell me what's slow, leaking, manual or guessed-at. I will tell you straight whether I can build something worth it, with a number attached. No long pitch, no obligation.

Email me See Scoutr
Melbourne, AustraliaRemote-friendly