Blog

What’s your role at Carbon, and what does your day-to-day involve?

As a developer at Carbon, I’m part of the team responsible for the company’s software, as well as its infrastructure, and our approach to innovation. 

My role involves coming up with new ideas, tackling industry challenges, and implementing practical, efficient solutions. Some days, that means developing new tools or refining existing ones; other days, it’s troubleshooting and problem-solving – and a bit of smashing my head against the keyboard when there’s a hard challenge to solve  –  but always refining and improving through testing and learning.

Having come from a background in finance, data science, and tech  –  insurance was completely new to me. I expected it to be slow-moving and traditional, but I’ve been pleasantly surprised to some extent particularly when it comes to culture – Carbon has a real culture of innovation and collaboration. However, one of the biggest adjustments has been dealing with insurance data. In finance, I was used to working with real-time, automated, and clean datasets. At Carbon, we need to process, clean, and structure the data ourselves, which presents an entirely different challenge.

What brought you to insurance? What were your preconceptions about the sector before joining?

My journey into insurance was somewhat unexpected. While finishing my master’s degree in data science, my CV was put forward to Mark Oldroyd, Carbon’s CTO, and I started working part-time. Initially, I was juggling my studies with two-days-a-week hybrid working in the office, where I worked alongside Carbon’s quant and development teams. It was intense but rewarding, giving me hands-on experience with cloud infrastructure and Artificial Intelligence (AI) in practice – building on the theory I was studying in parallel.

Before joining, I assumed insurance was quite old-fashioned, with slow processes and a reluctance to change. While that’s true in some areas, what I’ve learned is that real transformation is happening. The key challenge is balancing innovation with user habits – underwriters, for example, are often most comfortable working in spreadsheets. So, rather than completely overhauling workflows, it’s about nudging the industry forward in a way that integrates seamlessly with existing practices.

What are the biggest challenges in the sector right now?

When I first joined Carbon, I was struck by how complex the insurance market actually is – so many layers, so many moving parts. One of the biggest challenges is data quality. In insurance, we receive Excel bordereaux that are often missing key pieces of data or may have incorrect information in files therefore significant transformation and standardisation is required.

That’s where Graphene Studio comes in – a data preparation tool we’ve been developing at Carbon. Previously, our data operations team had to manually clean and input spreadsheets into our system. Now, 73% of bordereaux ingestion is automated through a no-code workflow tool, saving huge amounts of time. It also improves data quality, flagging errors like incorrect policy references or missing inception dates before they escalate into larger issues. Poor data quality isn’t just an administrative headache; in a worst-case scenario, it could snowball into significant losses.

How do tech and underwriting teams best work together?

Collaboration is crucial. At Carbon, we don’t operate in silos – our developers literally sit next to underwriters and operations teams, allowing for constant back-and-forth dialogue. This proximity means we see the challenges they face in real time and can build solutions that genuinely improve their day-to-day workflows.

A great example of this, is how Carbon uses AI to map the policy categories from coverholders into a smaller and more descriptive subset, which was subsequently finalised with the underwriting and claims teams. Now when new bordereaux is put into our system these categories are mapped automatically, making it easier for the teams to sort through the data.

What role do AI and automation have in insurance?

AI has huge potential, but its effectiveness depends on data quality. With clean data, AI can be transformative, spotting patterns humans might miss, improving predictive modeling, and even enhancing risk assessment.

One of the most promising applications is automation. AI can summarise documents, generate presentations, and handle administrative tasks that currently take up too much time. Instead of spending weeks compiling reports, AI can cut it down to days – or even hours. We’re still at the exploratory phase, but the industry should be looking at AI not as a threat but as a powerful tool for efficiency and accuracy.

Beyond automation, AI has significant potential to enhance predictive modeling and scenario modeling. By analysing vast datasets, AI can help underwriters simulate various risk scenarios with greater accuracy, improving decision-making and pricing strategies. This could be particularly valuable in areas like climate risk, where historical data alone is no longer sufficient to predict future trends.

How can the industry attract more young talent?

Tech professionals love solving problems – give us an interesting challenge, and we’ll jump at the opportunity. The insurance sector needs to tap into that mindset.

From my experience, insurance is doing a good job attracting young people, but not enough from tech backgrounds. That needs to change. There’s a misconception that insurance isn’t exciting, but in reality, it’s full of complex, rewarding challenges. Companies need to showcase these problems and invite tech talent to help solve them.

At Carbon, I’ve been given the space to be curious, to innovate, and to actively develop my skills for the tangible benefit of my team. If the industry can continue fostering that kind of environment, it won’t have trouble bringing in the next generation of talent.

GET IN TOUCH: If you’re interested in finding out more about Carbon, or would like a demo of our data platform Graphene, please reach out at info@carbonuw.com