Disclaimer: This article was AI generated from my YouTube video transcript.
Hi, I’m Blake, and I run Harkness AI in Christchurch. I’ve amassed over 1.4 million views on free AI educational content on social media, and I’ve had the opportunity to work with 25 unique businesses on their AI transformation and adoption. Along the way I noticed a huge gap for the younger generation, so I created Young Kiwis in AI - a community and platform to help get younger people into AI-related tech roles. And I’ve done all of this only a year and a half out of my uni degree. My goal is simple: help people and businesses work smarter, not harder, with AI.
The lightbulb moment
It came halfway through my university degree. I was studying mechatronic engineering, and around 2022 ChatGPT released. I chucked all of my exams and assignments through it and realised very quickly that it could answer everything I was meant to be learning. That made me think: am I actually going to get a role when I finish this degree? So I switched my focus purely to AI - learning everything about it and how I could help businesses actually utilise it.
The gap I found
The thing I noticed early on is that businesses are bombarded with flashy AI tools and demonstrations that don’t actually work in the real world. On the other side, there’s a huge number of AI developers who are deeply technical but can’t communicate and solve real business problems. That’s exactly where I found my niche: enough knowledge to build out AI systems, with the communication skills to relate them to real business problems - not just solving things for the sake of it.
How I actually work
I don’t start with the technology - I start with the team. We sit down together and figure out where the manual, repetitive tasks are that humans don’t actually provide real value on. That’s where AI and automation get focused. In practice that’s looked like building automations for invoicing and job reporting for a construction company, training teams internally on how to use their AI licences to improve their work, and more recently setting up Claude operating systems for smaller businesses - a 24/7 assistant built on their real files and information.
Here’s an example of the value that drives. A fencing supplier up in Auckland was drowning in admin - responding to emails, drafting and sending invoices, managing his calendar, even doing market research. He wasn’t able to spend time actually growing the business. Using Claude connected to his different tools, we’ve automated around 80-90% of that work. Now he just keeps a human review layer on top, and spends that time growing the business or with his wife and family.
And it’s not just small businesses. As a fractional AI engineer at Patersons - a land surveying and engineering firm with around 140 staff across the South Island - I’ve been developing the Patersons portal: around eight different applications covering resourcing, scheduling and data extraction, replacing the manual repetitive tasks our engineers and surveyors were doing.
Community impact
I bring the same approach to community. On my YouTube channel I post free, step-by-step educational content - for example, how to set Claude up with your Xero account. One of the coolest builds I’ve done is DigiKōr, a podcasting platform for young Māori and Pasifika students. They learn about podcasting, run interviews, use AI to help extract that information into a question-and-answer format - and it lets them earn NCEA credits along the way.
The other side of the impact is the younger generation. I get lots of people reaching out to say the content they’ve seen made them want to get into the AI space - they’ve started posting their own content and pushing forward in their careers. Through Young Kiwis in AI, we’ve also placed around five students into graduate or internship roles in the past five months.
The two rules everything runs on
Everything I build runs on two distinct rules: keep it explainable, and keep humans in the loop. AI is not a deterministic system - it won’t give you the same result every time you ask the same question. Because it’s probabilistic, we use it to draft results and keep human review and responsibility over the top. I instil this with every company I work with, and data always stays within the trusted infrastructure the business is already using.
I also put a big emphasis on the right tool for the job. Clients often come to me saying they need AI to solve a problem, when often plain automation or a script will do the job better.
I’ve accomplished a lot in a short time, but honestly, I’m just getting started - and I’d love to be New Zealand’s Rising AI Star for 2026. If you’d like to work smarter with AI in your own business, reach out for a free 30-minute chat.
