The VC firm’s got a fifth of the firm for their money (post-raise, Parsons is the largest investor with a 78% share, Icehouse owns 13%, Blackbird 7% and Sir Stephen Tindall’s K1W1 2%). Watchful also received a $400,000 matching R&D grant from the Crown-backed Callaghan Innovation to support its automation and AI projects.

While many founders create a firm to solve a problem they encountered in everyday life, Parsons’ example is at the sharp end of things.

He graduated from the Royal New Zealand Police College at just 17 (”My nickname was foetus”), with his swearing-in delayed a day until he turned 18. He then spent five years on the force through to 2018 – including a stint with the Tactical Crime Unit, AKA the burglary squad – in Papakura that saw him witness countless commercial and domestic break-ins. He got a first-hand look at the devastation, and also the tricks used to beat sometimes under-resourced security.

After leaving the police, he founded Crosbies Security Ltd (CSL), which began with home alarm installs and then branched out to develop mobile security towers – or solar-powered cameras on poles. Its first security tower contract was protecting Kāinga Ora developments during the Covid lockdowns. Clients including The Warehouse Group (where CSL was called in after a spate of ram raids), solar farm operator Lodestone Energy and KiwiRail followed.

Soon, CSL had more than 1000 security towers across 500 sites, and Parsons began to look for better ways to manage them. That led him to experiment with a large language model (LLM). Automated systems, including alerts, are old hat, come with even the cheapest cameras and have the basic smarts to, say, distinguish between a cat and a human.

But Parsons used generative AI to train Watchful to deal with much more complex causes of false positives. He says car yards are a classic. A system needs to gauge if a person lurking on the periphery is there to steal a catalytic converter – or a staffer working late, or a member of the public doing some window shopping.

Watchful uses AI to scour thousands of cameras for suspicious behaviour, then prioritise threats presented to human agents.

It’s a business-to-business play. The target market is security firms which monitor thousands or even tens of thousands of cameras for their various customers. The goal is for Watchful’s AI to automate as many actions previously taken by a human security agent as possible, eliminating errors and human bias and allowing agents to make swifter decisions.

“Watchful was an in-house project that I became addicted to,” Parsons says. He installed a general manager at CSL so he could concentrate on Watchful, which was spun out as a stand-alone company in December 2023.

In January, he made his first visits to customers during a trip to the US – the primary target for the Auckland-based Watchful simply because it’s the home to so many large scale surveillance firms. One prospect manages some 50,000 cameras across all 50 states, he says. “New Zealand is just too small for a return on investment.”

In good start-up tradition, his preliminary North American fact-finding mission was low-cost, as he house-swapped with a family in LA.

He was able to almost immediately sign customers, including Texas-based Zaladium Analytics, which runs a security monitoring center specialising in remote CCTV systems on retail and construction sites and a second Texan company, Eyeforce Remote Guarding Solutions.

By Parsons’ account, the Texan firm has realised a fourfold increase in efficiency as it uses Watchful’s companion product, Quill, to monitor thousands of security cameras at any one time, scouring live videos for potential threats.

These threats include high-risk events like break-ins, as well as more nuanced rules an organisation may set, such as a car tailgating an employee, or a person loitering next to a sensitive area, Parsons says.

“We have created a vastly improved triaging system, so instead of an agent trying to monitor hundreds of cameras at once, we can direct them to the most vital moments, to take decisive action, while automating the rest.”

The founder adds, “On my first trip to the United States with Watchful earlier this year, it was evident we had tapped into a significant global problem. Within weeks of on-boarding our first US client, the business was able to redeploy monitoring agents to other departments for work, which only a human could do.”

‘Ripe for AI-led disruption’

“Josh is a force of nature. Understanding that the speed of response is everything when aiming to deter a crime and Josh’s experience as a policeman and in building Crosbies allows him to think from first principles about how best to automate crime prevention,” Blackbird principal James Palmer says.

“Technology ultimately aims to deliver: Cheaper, faster and better. Watchful delivers on all three. We believe this category is ripe for AI-led disruption.”

“Weirdly enough, security surveillance is a market that hasn’t seen a huge amount of innovation in recent years. Recognising this, Joshua has figured out a way to address some of these clear gaps and has been rewarded with instant commercial traction,” says Mason Bleakley, the Icehouse Ventures principal who first approached the founder after his Hi-Tech Awards nomination.

Parsons says the $4.6m will go, in part, toward developing new features. Automated reporting is on the cards (which would take Watchful into the territory of Auror, the red-hot Auckland crime intelligence firm that recently raised $82m.

Meanwhile his original firm – now supercharged as a Watchful testbed customer – continues to hum. In 2024 alone, Crosbies Security triggered 555 police attendances, increased crime prevention to 85% of all intruders detected, and caught 44 arrests on camera, Parsons says. Its wins include two people prosecuted last month for the theft of $25,000 worth of aluminium.

“As the challenges around crime evolve, so too must the tools used to confront them. That’s why we are committed to continuous innovation, working alongside monitoring teams to ensure our technology remains at the cutting edge,” Parsons says.

Chris Keall is an Auckland-based member of the Herald’s business team. He joined the Herald in 2018 and is the technology editor and a senior business writer.