A delivery price calculator
Sales staff in store enter the delivery address and get a price based on distance from the store. Every customer is quoted the same way, and nobody guesses.
Automate the job somebody does by hand every Monday.
Marketing automation built with low-code tools and AI: small systems that take repetitive work off your team and connect the tools you already have. Built with checks, documented and handed over.
Based in Perth. Delivered remotely across Australia and overseas.
Why it matters.
Every marketing team has a few jobs that are done by hand, on a schedule, by someone who has better things to do: copying numbers between systems, updating a feed, compiling a report, checking competitor prices.
These are good candidates for automation, and AI has widened the list to include work that needs reading and judgement, such as sorting enquiries or drafting a first version. I build these as proper systems, with validation and a person approving anything that matters, because an automation that’s quietly wrong is worse than none.
This is for you if
- A recurring report takes someone half a day to assemble.
- Staff re-key the same details into more than one system.
- You’ve been quoted for a big software package to fix a small problem.
- Data is copied and pasted between tools.
- You want to use AI on real work and don’t know where to begin safely.
- You need outside data, such as competitor prices or listings, collected on a schedule.
Small systems, built as the need came up.
Four I built for one retailer. Each one could have been solved by buying a large piece of software. All four were built with low-code tools and AI inside an ongoing monthly retainer, as each problem came up.
A check on contact details
When a sale is entered, the system checks the email, phone number and address. If something looks misspelt or wrong, the store is emailed within 30 seconds to fix it while the customer is still there. That prevents the failed deliveries that lead to bad reviews.
A review request after delivery
Once an order is delivered, the buyer gets a text message asking them to review a locally owned retailer. It goes out by itself, at the moment they’re most likely to say yes.
Packing slips for the warehouse
Slips print automatically from the point of sale and inventory system, with the stock, the delivery details and any special instructions the warehouse needs.
None of these needed a new platform or a licence fee. They save staff time, cut mistakes that cost money later, and in the case of reviews, help bring in the next sale.
How the work runs.
Map
The task as it’s done today, step by step, including the exceptions.
Design
What to automate, what to leave to a person and where the approval points sit.
Build
Pipelines, integrations and AI steps using tools that suit your budget and skills.
Validate
Checks that catch bad data and failed runs, with alerts.
Hand over
Documentation and training, so your team can run and change it.
Deliverables
What you end up with
- Working automated workflow
- Validation and alerting
- Runbook
- Training
- Ownership of every account and licence
Where AI comes in
Faster, with a person checking
AI is often part of what gets built here. I use the same approach as in my own practice: written rules, your own data as the source of truth, and a person approving the output.
How I use AITools I work in
- Zapier
- Airtable
- Google Apps Script
- Google Sheets
- Python
- Anthropic Claude
- OpenRouter
Fair things to ask.
Something else? Ask me directly.
Do we have to buy new software?
Usually not. Most of what I build connects the tools you already pay for, such as your point of sale, your inbox and a spreadsheet, using low-code tools and a little AI. Where buying something is the better answer, I’ll say so.
Is each one a separate project?
It doesn’t have to be. Small systems like these fit well inside an ongoing monthly arrangement: a need comes up, I build the fix, and it’s covered by the hours already agreed.
Which tools do you use?
Whatever fits: Zapier or Airtable for simple jobs, Python and BigQuery for heavier ones, and AI models where the task involves reading or writing.
Is it safe to put our data through AI tools?
It depends on the data and the tool. I’ll set out what goes where before anything is built, and keep sensitive data out of services that don’t need it.
Can you automate our marketing reporting?
Usually, yes. Marketing reporting automation is one of the most common requests: a marketing data pipeline pulls the numbers from each platform on a schedule, checks them and fills the report, so nobody assembles it by hand. Marketing automation services cover the wider set of jobs beyond reporting.
Related work
AI
How I use AI
Systems with written rules, real data and approval points, in place of one-off prompts.
Read moreData
Dashboards and reporting
Looker Studio dashboards built on clean, reconciled sources and designed around the questions your team asks every month.
Read moreAdvertising
Dynamic & feed based ads
Dynamic product ads: an ad for every product you sell, built from your product feed, kept current when prices change, and shown to the people most likely to buy. Built, documented and handed over.
Read moreSomething driving you mad?
Tell me what’s going on. I’ll come back with questions or a scoped proposal, and I’ll say so if I’m not the right person for it.