Skip to main content


Insuretech Connect Asia 2024

26 Mar 2024

Navigating the Evolution: GenAI's Impact on the Future of Insurance Automation

Tricia Wong

This topic explores the intersection of genetic algorithms and artificial intelligence (GenAI) and its profound implications for the future of insurance automation. Delving into the innovative fusion of these technologies, the discussion elucidates how GenAI is poised to revolutionize insurance processes, from underwriting to claims management. By harnessing the power of genetic algorithms to optimize decision-making and predictive modeling, insurers can unlock new levels of efficiency, accuracy, and personalization.

We had the opportunity to interview David Huang, Chief Technology Officer at NEOS Life, where he shared insightful analysis and forward-thinking approaches regarding GenAI and the future of Insurance Automation.

In this interview, he addresses the transformative potential of artificial intelligence in revolutionizing insurance processes as well as the integration of advanced technologies like AI and machine learning to enhance customer experiences, streamline operations, and mitigate risks more effectively. By delving into the intricate dynamics of GenAI, David provides valuable insights into how NEOS Life and the wider insurance industry can adapt and thrive in an increasingly automated landscape, ultimately shaping the future of insurance services.

As we delve into the advancements in technology and artificial intelligence, particularly Generative Artificial Intelligence (GenAI), how do you see these innovations reshaping the landscape of the insurance industry?

GenAI is poised to revolutionize the insurance industry by automating complex tasks, improving risk assessment, and enhancing customer experiences. I believe we will see GenAI being applied across the insurance value chain, from insurance product development and underwriting to claims processing and customer service. By leveraging the power of machine learning and natural language processing, GenAI can help insurers make more accurate predictions, automate decision-making, and provide personalized services at scale. This will lead to greater efficiency, reduced costs, and improved customer satisfaction.

As Generative AI (GenAI) continues to evolve, its impact on the insurance industry's technological strategy and investment is both profound and transformative. The maturation of GenAI demands that insurers reevaluate their approach to technology investment, focusing on areas that will remain critical in the future. The advent of advanced models or the potential of Artificial General Intelligence (AGI) introduces a scenario where creating sophisticated front-end interfaces and robust back-end logic could become commodities, as simple as entering commands in natural language. This revolution raises important questions about the long-term value of current technology investments.

Insurance companies must consider how these advancements will affect not only their software development practices but also broader areas such as compliance controls, cybersecurity, business intelligence, and client reporting. The key to future-proofing technology strategy lies in recognizing the enduring importance of these domains, regardless of how AI capabilities evolve. Investments in technology must not only enhance current operational efficiencies but also provide the agility to adapt to future GenAI advancements. This approach ensures that insurers remain competitive and capable of leveraging GenAI to its fullest potential, thereby transforming challenges into opportunities for innovation and growth.

How is NEOS Life currently integrating GenAI into its operations, and what specific areas of insurance processes and services are being enhanced through this technology?

Firstly, we have adopted Retrieval-augmented generation (RAG) for our product and market research and comparisons. This approach allows us to synthesize vast amounts of data, thereby streamlining our market analysis and ensuring that our products are tailored to meet the evolving needs of our clientele. The application of RAG assists in distilling complex information into actionable insights, facilitating strategic decision-making and product development.

We also leverage existing business applications and keep train business users in different use cases and/or best practices. For instance, we educate internal staff to become better in prompt engineering, automate simple labor work, and be able to start generate valid ideas on GenAI use cases. Similarly, we introduced other open source LLM to show staff what is GenAI, LLM, why they are different and eventually how they are trained. These activities are great tool for us to build innovative culture through the company, make change management easier and be able to generate business value directly.

Furthermore, we are leveraging zero-shot learning in Large Language Models (LLMs) to expedite the underwriting pre-assessment process. This technology enables us to process applications faster than ever before, significantly reducing wait times and improving our overall service speed. By employing zero-shot learning, we can interpret and analyze customer information without prior explicit examples, leading to quicker and more accurate underwriting decisions.

Additionally, we are starting to utilize the integrated Generative AI (GenAI) capabilities of Amazon Connect to automate tasks such as call summarization, generating file notes, and creating future workflow tasks. These automated processes aim to enhance productivity and deliver an improved customer experience.

It is important to note that the incorporation of GenAI into our operations is not intended to replace our valued human workforce. Instead, our objective is to augment and empower our NEOS staff with the latest technological tools, enabling them to serve our customers more efficiently and effectively. Our AI projects are meticulously designed to maintain humans in the loop, thereby ensuring that our team's invaluable domain knowledge and experience are retained and enhanced.

By automating labor-intensive tasks such as data collection, filtering, and drafting, and by streamlining the review process against underwriting guidelines, we free up our staff to focus on the more nuanced and judgment-intensive aspects of their roles. This shift not only improves operational efficiency but also empowers our employees to make more informed decisions, leading to better outcomes for both NEOS Life and our customers.

In conclusion, NEOS Life is committed to harnessing the potential of GenAI to enhance our insurance processes and services. Through these advancements, we aim to improve efficiency, enhance customer satisfaction, and empower our staff, thereby solidifying our position as a leading, innovative provider in the insurance industry. Our dedication to integrating advanced technologies, coupled with our commitment to our employees and clients, ensures that NEOS Life remains at the cutting edge of the insurance sector.

What were some of the challenges encountered in the adoption of GenAI within the insurance domain?

Integrating Generative AI (GenAI) into the insurance sector is an exciting yet challenging journey. At NEOS Life, we've navigated these waters with a strategic blend of innovation, responsibility, and transparency.

Design the AI products together: There are lessons learned to co-design AI products together with business teams at early stages. This can increase in identifying and framing the correct business problem, uncovering hidden knowledge, and ensure the AI product can generate business value quickly.

Overcoming Technical Hurdles: Data privacy sits at the core of our GenAI adoption strategy. We've fortified our systems against breaches, ensuring our customers' data remains secure and confidential. All GenAI is using private infrastructure within AWS rather than using public API to ChatGPT  or Mixtral directly. To addressing potential biases in AI, we've committed to using human labelled dataset to fine-tuning the LLM models and set compliance guardrails around them to ensure fairness and accuracy in all AI-generated insights. We also implemented LangSmith and other tracking tools to keep monitoring the response quality.

Building Trust: Perhaps the most crucial aspect of our journey has been in winning over our employees. We understand the importance of the human touch in insurance. By educating our key staff on the benefits of GenAI and ensuring human oversight in critical areas, we've started to turn skepticism into optimism. Continuous training and upskilling ensure they feel confident and capable in this new era of insurance, bridging any skills gaps and fostering a culture of innovation.

In conclusion, the path to integrating GenAI in insurance is complex, but at NEOS Life, we're tackling these challenges head-on. Through strategic investments in technology, a keen eye on regulation, and a commitment to our customers and employees, we're not just adapting to a new era of insurance—we're leading it.

With the growing reliance on data for AI-driven decision-making, how does NEOS Life address concerns related to data security and privacy, ensuring that customer information is safeguarded while still harnessing the power of AI for improved insurance processes?

At NEOS Life, we deeply understand the critical importance of data security and privacy in today’s digital age, especially when leveraging AI for insurance processes. Our approach is holistic, focusing on both protecting our customers’ personal information and utilizing AI to enhance our services effectively.

Advanced Encryption and Cybersecurity Measures: We employ state-of-the-art encryption techniques to secure our data at rest and in transit. Our cybersecurity infrastructure on AWS is robust, incorporating the latest in threat detection and mitigation technologies to guard against unauthorized access and potential data breaches.

Transparent AI Operations: Transparency is key in our AI-driven processes. We ensure our AI systems are explainable and decisions made by AI are understandable to our customers. For example, by using open source project from LangChain and Llamaindex, we have built lots of observabilities into the product development, training and production cycles.

Synthetic Data first approach: When possible, we use existing LLM to generate synthetic data as much as possible. We call this “Synthetic Data First” approach. Using this approach, not only we can keep capture hidden business rules, but also improve the AI decision and result quality through human feedback in the testing. This approach is unique and helped us quickly and safely put together GenAI products.

Continuous Improvement and Training: The landscape of AI, data security, and privacy is constantly evolving. NEOS Life invests in continuous training for our staff and regularly updates our policies and technologies to stay ahead of emerging threats and ensure our practices reflect the latest standards in data protection.

Looking ahead, what emerging trends and innovations do you anticipate in the intersection of insurance, technology, and artificial intelligence?

As we peer into the future of insurance, technology, and artificial intelligence, we see a landscape ripe with potential for groundbreaking innovations and transformative trends. At NEOS Life, our commitment to staying ahead of the curve drives us to closely monitor and engage with these developments.

Self serve GenAI product building process: with more power user from the existing GenAI products, we are getting more and more useful feedback and potential use cases. In addition, GenAI tooling is getting more mature over time. With strong data capability and modern data lake, we are aiming to make this process more self serve by the business users. So that our business users are pairing up with our software and data engineers to automate more business processes quicker, from data discovery, ideation, prototyping all the way to productionalise.

Better Synthetic data generation: The future of synthetic data generation in the insurance industry heralds a transformative era for risk assessment, fraud detection, and personalized customer experiences. As we venture forward, the evolution of more sophisticated and accurate synthetic data algorithms stands to revolutionize how insurers understand and interact with their customers. This leap in technology promises to generate high-quality, anonymized datasets that mirror real-world complexities without compromising individual privacy. Such advancements will enable insurers to refine their models with a breadth of data previously inaccessible, driving innovation in product offerings, pricing strategies, and customer service. For the insurance industry, better synthetic data generation means not just enhanced operational efficiency but also a significant step toward more ethical and responsible data use, ensuring customer trust and regulatory compliance in an increasingly data-driven world.

Evolution of AI Algorithms for Personalized Insurance: AI's role in insurance is set to deepen, with algorithms becoming more sophisticated in predicting risks and customizing solutions. This will enable us to offer highly personalized insurance products, tailored to the unique needs and lifestyles of our customers. By harnessing advanced data analytics and machine learning, we aim to not only anticipate our customers' needs but also deliver services that are more relevant and valuable.

In conclusion, the future of insurance at the intersection with technology and AI is bright with possibilities. At NEOS Life, we're not just observing these trends; we're actively participating in shaping them. By embracing innovation, ensuring the ethical use of technology, and focusing on personalized and efficient customer service, we're paving the way for a new era in the insurance industry.

View all Articles