Artificial intelligence firm Imagr is hoping its tech will change the way society functions. Founder and former assistant portfolio fund manager William Chomley talks business motivation and what the future looks like for AI.
A brief description of the business
Imagr is an Auckland based AI company that specialises in image processing and recognition. This means through our software we are able to teach a computer to act like a human eye. Our keystone product is Smartcart, a trolley-mounted camera that removes the need for checkouts in supermarkets. I started the business in 2015, in Sydney, but started operating here last July. We have 10 employees.
What sparked the idea?
The Smartcart concept was born out of frustration with the retail sector. I used to work above a Woolworths supermarket and would spend the majority of my lunch break waiting in line. The fact that we still have to line up at a checkout is archaic. I was researching the change in other industries and the effect AI was having and decided to jump in head first.
Yes, my parents moved us to Sydney when I was 16, so I did my last year of school there.
I've always wanted to move back to Auckland, but needed a catalyst and this was it. Imagr started as a result of a failed attempt at Smartcart, using a technology called RFID. It wasn't until I got involved in AI, and started a funding round that was filled by a majority of NZ investors, that I decided it was my time to move back.
As we have a unique base algorithm we charge our customers a customisation fee, this allows us to do on-barding and re-work our platform to understand the objects we need to be looking at for clients and the integration of our platform into their current systems. Once up and running we have a "software as a service" model where we charge a flat fee per transaction. For example, every time someone connects their phone to a trolley or to a basket we take a fixed fee.
What brands are you working with?
We're working alongside some of the largest retailers in the world. As the next steps are still in discussion we're unfortunately not able to publicly name them.
HuaJie Ni, Howard Geng, Joshua Jordan, Konstantinos Raptis, Mohammad Hedayati, Sonja Ridder.
What's it like running an artificial intelligence firm?
I think it has a few additional hurdles than running a normal software firm. The biggest hurdle is that the tech is still embryonic, so there's a larger knowledge gap to overcome for people to want to adopt it. At the same time every day is so exciting - we're at the forefront of change in a market, with software that has no manual to follow. I work with some of the greatest minds in the world and never has being "the dumbest in the room" been so true as it is with my case.
How are you learning the skills to be able to develop AI?
My background is in finance, so it's been a huge learning curve for me. I sit with my team for a few hours each week to understand the technology aspect. Coursera has also been amazing and most of the peer reviewed articles that my team read are passed on. I'm very fortunate to be surrounded by the team I am because they've been fantastic at bringing me up to speed.
Surround yourself with brilliant people who like to solve challenges because AI is a round which is a lot less travelled - there's no manual for how to do things.
How hard is it to find specialist staff here in New Zealand?
It's hard to find skilled talent here in New Zealand, for a few reasons. The first is that there are only five AI companies here, so jobs are scarce. Anyone that finishes with a degree in this field is most likely going to head overseas to find work. It's a new topic in the area of computer science, so anyone that is at masters level or above would have had to have been involved five to 10 years ago in this field, so the pool is quite small. I've been lucky enough to source two computer vision engineers from New Zealand, we imported one other from the US.
AI is still relatively small, both here and in Australia. The AI market will have an effect on every industry and every sector. The ability for this form of computer science to streamline inefficiency is unparalleled. Annual worldwide AI revenue will grow from $643.7 million in 2016 to $38.8 billion by 2025.
William Chomley, founder of Imagr.
What are your long-term plans for Imagr?
By the start of next year we believe we'll be in a small format New Zealand store with our product Smartcart, and to do this our team will need to almost double. Long term, we see a lot of benefit with sticking to the retail sector and within five years we hope to be across two retailers, both here and internationally. We believe our technology can be applied to other sectors and will allow for streamlining in gas and utility, fuel retail and even facial recognition to assist with smart buildings.
What advice do you give to others thinking of starting a business?
If you feel there's an area of automation that could be achieved in an industry, then AI can probably help you streamline this.
There's a lot of untapped opportunity out there so go for it. Try to stay away from using too much open source software. Building from scratch, although it's harder, will prove to be more fruitful in the long run as you build a lot more initial IP and aren't held ransom by other companies' software. Surround yourself with brilliant people who like to solve challenges because AI is a road which is a lot less travelled - there's no manual for how to do things so having a strong team is really important.
