With predictive underwriting models, Digit excels in granular risk segmentation, ensuring accurate pricing and profitability. The company’s tech-driven approach includes AI bots for seamless operations, sentiment analysis for customer calls, and an intent-classification AI for efficient email handling, shared Vishal Shah, Head – Data Science, Digit General Insurance, in an exclusive interaction with Srajan Agarwal of Elets News Network (ENN).
As the Head of Data Science, can you share some insights into how you leverage data analytics to enhance decisionmaking processes within the company, particularly in the context of risk assessment and underwriting?
As an insurance company, we use various data points when underwriting insurance for a new customer, and data science plays a crucial role in this. We have harnessed our tech platform to develop predictive underwriting models that leverage the insights gathered by our data bank. This aids us in in determining and targeting the markets and customers in India that are expected to be more profitable and hence, allows us to accurately price our coverage. Underwriting models are inherently predictive since underwriting models try to predict future outcomes. However, our models are automated.
We use rule engines and granular-level risk segmentation by using more variables and a higher level of granularity. We believe that this differentiates us from other underwriting models. Our models are used to improve risk selection and identify low-risk customers as compared to high-risk customers. For example, in motor insurance, we use variables including fuel, vehicle make, model and subtype, odometer reading, vehicle location, coverages opted, sourcing channel, past claim history, and usage of vehicles to improve the granularity of risk selection using predictive underwriting models. This enables us to charge an appropriate price to each customer, thereby lowering the loss ratio. It provides us with an advantage in acquiring and retaining target customers while avoiding underwriting less profitable business.
We believe this has helped us overcome the challenges historically associated with underwriting motor insurance in India, allowing us to capitalise on the sizable opportunity in our market. We expect our growth to give rise to a feedback loop by driving the accumulation of more data, which will aid the predictive power of our models, further strengthening these advantages and ultimately resulting in profitable growth.
In the rapidly evolving landscape of insurtech, how has Digit General Insurance embraced innovative data science methodologies to stay ahead of the curve? Can you highlight specific projects or technologies that have significantly impacted the company’s approach to insurance?
Our use of technology and AI-driven microsystems, or bots, to streamline a major amount of our operations across the onboarding, underwriting, servicing, and claims processes has allowed us to deliver a high-quality customer and partner experience while keeping our employee base lean. As we have become pioneers in customer satisfaction through our mission of simplifying insurance, our efforts are now moving towards customer delight to provide them with a more nuanced and heightened experience. Tech is at the core of everything that we do at Digit, and we have come up with various AI-ML projects to change how insurance is serviced in India. For example, we have developed an AI-based sentiment analysis model to detect sentiment in audio and text, analyse customer calls in real-time, and prompt the team leader intervention where the score is a high negative. We have also built an intent-classification AI engine that reads incoming emails and predicts their intent, assigns priority and auto-generates replies through robotic automation, thus reducing response time and improving efficiency.
With the increasing importance of AI in the insurance industry, what role do you see data science playing in the future of insurance, and how is Digit General Insurance preparing for the upcoming advancements in AI and ML?
Technology and digital analytics are at the core of our business. As a relatively young general insurance company, we designed our digital infrastructure from inception with a focus on delivering a high-quality customer experience. Our strong execution track record underpinned by this nimble model is evidenced by our ability to identify market opportunities, balance organic growth, and speedily activate new partnerships and integration. For instance, in March 2020, amid the COVID-19 pandemic, we were one of the first insurers in India to launch a customisable group illness insurance product covering COVID-19 hospitalisation costs, aimed at protecting employees and distributors of small- to medium-sized enterprises from medical expenses that could arise as a result of infection.
We build technology-enabled solutions and employ a hybrid model of AI-enabled analytics and human assessment to streamline the value chain, aid our customers, partners, and employees, and drive efficiency. Around the core of our technology platform, we have developed in-house microsystems, that allow us to facilitate a range of routine tasks, from policy design, underwriting, pricing, and issuance to servicing and claims management. Our platform is entirely cloud-based, making our system agile, connected, and scalable. We utilise AI and machine learning to enhance efficiency and have 459 active AIdriven microsystems to automate processes for the benefit of our partners and customers. As a digital-first general insurance company, the majority of the data we have collected has been used to teach our microsystems to evaluate the applications and claims we receive. We have developed self-service options with 24×7 live chatbot assistance for our customers and partners on popular messaging tools such as WhatsApp, as well as on our website.
Given the ever-growing concerns around data privacy and security, how does Digit General Insurance address these issues in the development and implementation of data science models? What measures are in place to safeguard customer information while extracting meaningful insights?
We have implemented internal policies regarding IT and data security, data privacy as well as responses to data subject access rights. These policies and their implementation are regularly reviewed and audited by a dedicated team of information security professionals. Our privacy policy specifies the framework for proactive threat detection and prevention, ensuring integrity and validity of data contained in information systems, consistent and secure use of information, efficient and effective recovery from information system disruption, and protection of our IT assets, including information, software, configurations and hardware. Further, we have comprehensive programs on responsible disclosures and vulnerability management. Our information security team, along with third-party specialists, conducts regular security assessments and penetration tests on our applications, cloud infrastructure, workstations, and network equipment, following which remedial measures are implemented where necessary.
We use web application firewalls and customised solutions as defensive mechanisms against malicious traffic, hacking, and distributed denial of service attempts and encrypt all data during transit using strong cryptographic protocols. We use multi-factor authentication and other security controls to control access to, and authorised use of, personal data or other confidential information.
Elets The Banking and Finance Post Magazine has carved out a niche for itself in the crowded market with exclusive & unique content. Get in-depth insights on trend-setting innovations & transformation in the BFSI sector. Best offers for Print + Digital issues! Subscribe here➔ www.eletsonline.com/subscription/
