I co-founded and ran Sky-Futures as the CEO from 2009 to 2019 when I sold the business to private equity. We built a global business providing managed drone inspection services and inspection and visualisation software. Along the way, we counted some of the biggest companies on the planet as customers and top Venture Capital and Corporate Investors.

It was an intense learning experience, both for me personally and my team. We experienced hyper-growth, hiring at all levels from drone pilots to executives and managing a global team across multiple time zones.

Same trusted team, new opportunity

The decision to sell the business and spin out the software was not borne out of a desire to sit on a beach. My passion is solving business problems, using innovative technology and business models that go against the status quo and achieve a lasting and positive environmental impact. It stemmed from a belief that we had a massive opportunity with the technology and team that trusted and came with me. As for what that looked like in May 2019 on day 1, we didn’t know. We had a small team, and I was back as a ‘pre-seed’ Founder. We had relatively short feedback loops from our customers into product development, and I got my hands dirty once again covering all the non-engineering business functions.

In truth, I am the only non-technical function in the business. But I deeply intertwine with the product strategy, wearing the hats of our current and future customers, and continually challenge the team to make the complex simple. In this role, I’m provocative, I antagonise and frustrate because customer simplicity is not easy. But it is a fundamental guiding principle of this business.

Constant innovation drives our product development

We started with an inspection and visualisation platform. Inspection² application provides semi-automated and fully automatic asset inspection for the industry. We successfully service large, multi-national clients across a range of sectors with our solution today. A key consideration for the business is the length of Enterprise B2B buying cycles. We knew that we had a bigger and wider opportunity but had more work ahead to bring it to life.

Part of the attraction of working in a start-up is the ability to push boundaries – delivering technically challenging and interesting projects. We let our team work on anything they want for half a day a week. This idea doesn’t have to solve an immediate problem, it can be anything. One of the ideas we pursued was to try and use Image based AI to automate anomaly detection. It was a side project initiated in 2016 by Thomas, our head of AI, using state vector machines. It continued in fits and starts continually trialling cutting-edge tech such as convolutional deep neural networks all the way through to the beginning of the pandemic.

When the pandemic hit, we knew that our big corporate customers and new opportunities would slow down. We sensed that the timing was right, so we kept pushing, innovating, and exploring to find the right solution for the customer, which would allow us to have a greater impact across sectors. With a major refactor of the platform already planned, we began in earnest.

As we were assessing the project, cautious optimism emerged. Thomas had started to build tools to make the Image-based AI work quicker and more accurately.

We were approached by a large enterprise customer in Transmission and Distribution before the pandemic and were selected from a large pool of potential partners based on our technical expertise. We discussed the problem and mutually decided to set and take on a very difficult technical challenge. The results are excellent, and we will be sharing them publicly with our customers soon.

Democratising image AI for everyone

We still had a problem. As a business, we didn’t want to build a classic long cycle, customer feature-driven B2B business. We are here to make an impact on the universe, otherwise, we’d be working for big tech. We needed a quick assessment of product/market fit, so I started to explore use cases and approached two industries, construction and agriculture.

The reception was positive, as we explained our deeply technical heritage in developing Image AI tools and our relentless focus on customer experience. We can build AI models that solve some of their most pressing issues, at scale. In just eight weeks, we now have two PoCs lined up, with the respective sector leaders. But what’s better is that we have also created a B2B sales approach that circumnavigates the classic B2B sales cycle.

It takes time for each industry’s first PoC as we navigate the industry itself to find the most pressing and repeatable problems to solve at scale. Once we understand the most common scalable issues we can prebuild AI models for these business scenarios, ready to be deployed in weeks with a simple tuning to optimise each AI model’s performance.

Decisions can be made quickly, and technology can be implemented in weeks. Our focus is scale, scale, scale.

We spun out an amazing team that has worked together for an average of 6 years. I am deeply grateful to them all, they are truly excellent. We have a strong base of clients actively using our visualisation and inspection platform. We also have deep expertise in the problem set. Focus now turns to build the scalable tech and business model that is democratising Image AI for everyone. With broad, cross-sector applications, it will also help to solve problems that can help combat climate change. For us, once again it’s just the beginning of the journey, but with our heritage, we have the ability to make a significant impact on the universe.

Inspection² is particularly interested in solving problems that can help combat climate change and global warming, for example, analysing satellite data for methane clouds or tracking algae blooms on rivers. If you have a problem that you think AI Models can help with, please get in touch.

James Harison

James Harison

Founder & CEO

Pin It on Pinterest

Share This