Interview with Arina Man, CEO K2G, Given to XPRIMM

XPRIMM: K2G positions itself as a pioneer in AI-powered visual intelligence for the insurance industry. How did the company come to focus on claims automation?

Arina MAN: K2G focused on claims automation because this is exactly what our clients have been asking for. Across the insurance industry, companies are actively looking for ways to move away from slow, repetitive manual work. They need faster, more accurate decisions and a smoother customer experience.

As we worked with insurers, we kept hearing the same challenge: claims take too much time, and people are overloaded with routine tasks. So we built AI-powered visual intelligence solutions that directly address this demand. Claims automation became a natural focus because it brings immediate value - higher speed, better accuracy, and more efficient operations for our clients.

XPRIMM: Your technology delivers instant damage assessment through AI. What do you see as the most transformative impact of these tools on the motor claims workflow?

A.M.: Motor claims benefit from AI more than almost any other line of business because they combine three things: high volume, high complexity, and high customer expectations. Each claim involves multiple documents, photos, repair standards, deadlines, and interactions with external partners. Even small delays or inconsistencies quickly accumulate into higher costs and lower satisfaction.

K2G Claim Management AI Agent is particularly powerful here because it brings structure and speed to a workflow that is traditionally fragmented. In motor claims, where repair costs are volatile and cases often evolve quickly, this level of automation reduces leakage, improves accuracy, and prevents delays that could lead to penalties.

XPRIMM: Accuracy and trust are essential for automated claims. How does K2G ensure the reliability of its AI models, especially considering the diversity of vehicles, repair standards and local market practices? 

A.M.: Accuracy and trust are absolutely fundamental for automated claims, and we design our technology with that mindset from the very beginning. In motor insurance specifically, ensuring reliability means focusing both on how our models learn and how our models learn and how they are validated.

Reliability is only one side of trust. Equally important is transparency. Our platform not only generates an answer but also shows the logic behind it and the reasoning path the model followed. Users see why the AI made a decision and what calculations were behind it, which builds confidence rather than dependency.

To further strengthen this, our QA Agent runs a secondary verification. It recalculates the result using an independent piece of code to catch potential errors or hallucinations - essentially providing a built-in second opinion.

XPRIMM: Insurers often face legacy IT systems and fragmented data. What integration or change-management challenges do you observe when onboarding new clients, and how do you help them overcome these barriers?

A.M.: Our agents are built to be highly adaptable and work with any core system.

Fragmented data is another common pain point, and in fact, one of our favourite problems to solve. Our K2G Cleansing Agent handles this exceptionally well by connecting, structuring, and interpreting information from multiple sources. This allows insurers to move forward with AI even if their data landscape is far from perfect.

Understandably, legacy IT systems are naturally a concern for many insurers. Any system begins to age the minute it’s released. That’s the reality of working in a fast-moving technological world. My advice for insurers: to choose core systems that deliver regular updates. System upgrades can be painful, but it’s far better for the CTO to manage planned improvements than to allow legacy systems to grow like a snowball.

XPRIMM: As electric and connected vehicles become more common, repair complexity and cost variability are increasing. How is K2G adapting its technology to reflect these new realities?

A.M.: The core challenge with electric and connected vehicles is simple: insurers must underwrite and handle claims today, but the statistical data needed to model these risks accurately remains limited. At the same time, the volume of this data is growing rapidly - every month the industry learns more about real repair costs, failure patterns, parts availability, and the impact of new technologies.

K2G AI Agents (Reporting Agent and Talk-to-Data Agent) can deliver insights in minutes that would normally take weeks of manual analysis. This speed allows insurers to make confident, data-driven decisions even in areas where historical datasets are still limited but expanding quickly.

XPRIMM: Fraud remains a major concern in motor insurance. How can AI-based image analytics and behavioural patterns help insurers identify high-risk cases without disrupting the customer experience?

A.M.: Fraud is a serious and constantly evolving challenge in motor insurance. It doesn’t appear in isolation - it often emerges as a reaction to regulatory changes, economic pressures, or even geopolitical events. Insurers have become much better at dealing with known fraud patterns, so today the focus is shifting toward monitoring new schemes and strengthening prevention rather than reacting after the fact.

At the same time, we shouldn’t limit the conversation to fraud alone. There are growing sources of potential losses - higher repair costs, the impact of new vehicle technologies, and increasing regulatory and consumer expectations. All of these factors directly influence the customer experience.

Modern customers expect a personalised, seamless, digital process when buying insurance, managing their policy, or filing a claim. To deliver that, insurers must either heavily invest in proprietary technology or partner with agile InsurTech companies. AI allows insurers to maintain strong protection while keeping the customer experience fast, simple, and frictionless.

XPRIMM: Looking ahead, what strategic priorities define K2G’s roadmap for the next few years? Are you focusing on new lines of business, broader geographic expansion, or deeper automation across the claims value chain?

A.M.: That’s a great question. As the CEO, long-term strategic planning is one of my core responsibilities - and it’s something our shareholders expect from me as well. For K2G, our long-term direction always goes back to our core values: innovation, benefit, and profit. In practice, this means we closely monitor emerging technologies and market trends, but we only adopt what genuinely brings value to our clients, to society, and ultimately to our shareholders.

At the same time, we continuously review and adjust our product strategy. When you work with cutting-edge technologies, new capabilities appear all the time, and staying agile is essential. We prefer to evolve quickly rather than lock ourselves into a rigid multi-year plan.

What does this mean for our clients? It means constant monitoring of market needs and client feedback, fast iteration, and rapid release of new agents. It means deepening and strengthening our existing products while exploring new ones in parallel. And it also means building strong strategic partnerships with other innovators - for example, our collaboration with Revauxy, a leader in insurance claims administration

So instead of choosing between new business lines, geographic expansion, or deeper automation, our roadmap combines all three - but always through the lens of delivering real, meaningful value.

Original on Xprimm.com