The latest instalment of the Pearson’s Skill Outlook series claimed that UK workers could collectively save 19 million hours a week by 2026 by using generative AI for routine and repetitive tasks.

Based on their research, Pearson found that generative AI can most effectively support tasks related to record maintenance, data collection, or researching and compiling information for others.

There’s no shortage of data out there vaunting GenAI as we play witness to the opening of a technological frontier in which everyone’s scrambling for position at the front of the queue.

One thing that AI can inarguably do well is process data to deliver insight in a way that’s accessible to pretty much everyone – on the proviso that you give it good data if you want good insight on the other side.

It’s with this in mind that using AI for organisational insight and productivity benefits are among the most arguably intuitive uses for the tech out there in terms of delivering tangibly positive business outcomes using data that you – more than likely – already have access to.

One Scottish startup making waves in this space is Columbus Tech, who’s platform seeks to boost organisational performance through AI insights, strategy execution automation, and a better employee experience.

We spoke to the company’s managing director, Kat Gref, to have a candid discussion around the potential benefits, concerning pitfalls, and useful applications at this nascent stage for the technology, while also talking about how Columbus Tech is taking its place in this new frontier.

Empowering Scottish Startups with AI

Deriving insight from data is nothing new. It’s been an imperative for business success for years now.

The problem is, any kind of compelling “observability” was not really accessible for smaller organisations in the past. It was complex and wildly expensive, with algorithms being built from scratch to connect the data and the dots from different systems.

On this, Gref says: “Traditionally, implementing business intelligence solutions required significant financial investment and technical expertise.”

So, the technology is here, it’s accessible (depending on what you’re trying to achieve and at what scale) and it’s undoubtedly everywhere in the discourse of business and tech.

But why? If you’re a Scottish startup with a handful of employees working in financial services (for example) – why do you need this tech and what do you use it for?

According to Gref, using AI to drive observability and all the good stuff that comes with it, from an strategy execution and managing performance point of view, isn’t just for big organisations.

She says: “[AI] can be used by anybody on the scale of operation size. We’re a startup ourselves, using the system, to automate business reviews and generate performance insights. It keeps our team streamlined, informed and focused on higher-value work.

“Specifically, within the business with the idea of you need to have very specific goals, and your staff – no matter how many of them there are – need to have that clarity of where you’re going.

“If you can do this – and AI can help with that immensely – it can make your operation agile and dynamic, an important quality when you consider the pace of change.”

Essentially, using an AI-powered productivity/data insight platform gives you “a very simple tool to make sure your organisation is aligned on what you’re trying to achieve.”

While that certainly represents the best of GenAI’s potential, Gref isn’t naïve to the many dangers that the technology presents and is candid in her assessment of where we’re at when it comes to responsibly integrating the technology.

AI’s Impact on Corporate Decision-Making

A recent IBM study that showed that 64% of CEOs are adopting generative AI faster than staff are comfortable with in order to stay ahead of the competition.

Discussing how businesses can ease the transition to AI-led processes, Gref says: “We’re all still very much in an experimental stage right now. I think the potential is great, but keeping that human feedback is very important.

“There are inherent risks, including biases and ethical dilemmas.

“AI can produce information that can be very, very convincing, right? You put something in and it comes back with information in a good, presentable format. And that – on the face of it –  is very credible.

“However, especially in the corporate space, we need to be very mindful that this information can be incorrect, it can be biased and it can be unethical.”

This is why platform’s like Columbus Tech operate using a walled garden of data generated by your business so as not to rely on the dice roll of truth that’s often associated with GenAI LLM’s.

According to Gref, Columbus Tech’s platform leverages “industry frameworks and domain knowledge from working with over 100 organisations worldwide.”

She adds: “We organise business data according to these frameworks to deliver tailored AI insights. Our models are continuously refined through human oversight, customised for client needs, and integrated into enterprise landscapes.

“This enhances decision-making by collecting and analysing performance metrics, such as financial, customer, and project data.

Speaking more broadly about the unfolding proliferation of AI, Gref feels that it’s problematic that there’s a distinct lack of rigorous frameworks which gives businesses clear guidelines within which to innovate – as it stands, businesses don’t know if what they’re doing will be compliant or not down the line.

Gref’s point is valid one, when you consider a recent report from The AI Safety Institute (AISI), which showed that despite rules designed to prevent misuse, the largest publicly available language models (LLMs) can be easily manipulated to provide inappropriate or harmful answers.

This issue compounds when you consider that “with Gen AI models, once the data goes in, it’s almost impossible to trace back,” says Gref.

Scaling Challenges for AI-focused Startups

Despite the widespread recognition of AI’s importance, Gref highlighted a significant gap in adoption, stating, “Only about 15% of strategic planning, execution, and reporting are currently being automated.”

It’s with this in mind that we discuss the challenges of a Scottish business entering a global market that’s fiercely competitive.

As a startup, particularly those like Columbus Tech that focuses on AI, transitioning from initial development to the scaling phase, Gref says they face two primary challenges: securing necessary investments and expanding market reach beyond early adopters.

Recommended reading

• New Agreement Aims to Boost Scotland’s Productivity and Innovation
• Gov Plans to Make UK ‘International Tech Superpower’ Revealed
• AWS: UK on Track to be an AI Superpower

The investment dilemma often becomes a catch-22, where financing is needed to develop a competitive product and gain initial customers, yet these elements are also prerequisites for attracting investment.

Once past the early stages, the next hurdle is scaling the business, which requires a different set of skills focused on market entry and commercialisation.

On this, Gref says: “Despite having a solid product market fit and finances, many organisations, including high-profile unicorns like WeWork and Theranos, fail at this critical stage, where they struggle to commercialise their propositions and scale effectively.

“This highlights the crucial need for an ecosystem that supports businesses in their growth and scale-up stages, ensuring they have the tools and guidance necessary to navigate these wobbly transitions successfully.”