Feilding coder aims to take over the world one stockroom at a time

Feilding coder Paul Simpson has spent years developing a "machine learning" algorithm that predicts the amount of stock ...
DAVID UNWIN/STUFF

Feilding coder Paul Simpson has spent years developing a "machine learning" algorithm that predicts the amount of stock a business will need.

A  Feilding computer programmer plans to "take over the world" with his artificial intelligence invention.

Paul Simpson has spent four years perfecting his "machine learning algorithm" that predicts how much stock a company needs to have on hand.

It's right on the forefront of software development – there's nothing like it on the market in New Zealand and it has less than a handful of rivals overseas, Simpson said.

Before his creation, called StockTrim, predictive inventory management software took months to customise and install, and could cost hundreds of thousands of dollars  upfront. Simpson's system runs online, takes minutes to set up and costs a small monthly subscription.

READ MORE:
* Family tradgedy inspires entrepreneur to build online 'life-raft' for mourners
* Innovative inventors ready for final face-off
* Ōtaki entrepreneur's award-winning wheelchair
* Award-winning agricultural app developer hopes to tap Central Field Days crowds

"The aim is to take over the world – to do the same thing Xero has done for small business accounting," Simpson said.

His idea has already earned him a place as a finalist in Innovate 2017, an entrepreneurs' contest run in Manawatū.

Innovate mentor Neriah Broughton said Simpson already had his product and a fledgling business in StockTrim, with two or three clients under his belt. He just needed some support to really get going.

That made him an ideal Innovate candidate, but it was his algorithm's wider applications, from risk management to city planning, that made Broughton eager to be his main mentor.

"When there's an idea like Paul's, I get excited. It's like I can see the future."

Ad Feedback

The current focus was on perfecting it as an inventory-management aid. But the program also works for tasks involving predictions based on lots of information, Simpson said.

For example, in city planning, councils have information on population trends, where the current infrastructure is found and how well it meets demand.

Simpson's algorithm could use that to predict where the best places are to put drains, sewers and power lines. And the more data and feedback it gets, the better its predictions become, he said.

Broughton will guide Simpson through Innovate's 10-week crash course on business development, to help him prepare his pitch for the final in November.

 - Stuff

Comments

Ad Feedback
special offers
Ad Feedback