Amidst the AI revolution, e27 presents a new series showcasing how organisations embrace AI in their operations.

Since 2017, Suradej Panich has led the Data Scientist team at Sunday, a group company providing integrated data and technology platforms, including real-time customised non-life insurance, healthcare services, and a technology hub in Thailand.

Panich’s projects are central to Sunday’s core business, spanning the complete value chain. Additionally, he oversees vital partnerships with prominent companies across Southeast Asian industries.

In this edition, Panich shares how Sunday has embraced Artificial Intelligence.

Edited excerpts:

How do you perceive the AI revolution and its potential impact on your industry and workforce?

I think it’s no longer a question for company executives on the impact of Artificial Intelligence. Over the past ten years, many firms have moved from questioning to adopting or nurturing the technology to sit at the core differentiations of their products and services.

Here are some of the ways that AI is already being used in Sunday insurance and how it could impact the industry in the future:

Automating tasks: AI can automate many tasks currently performed by humans in the insurance industry, such as data entry, claims processing, and customer service. This can free human employees to focus on more complex and value-added tasks, such as risk assessment and underwriting.
Personalising products and services: AI can collect and analyse vast amounts of customer data, which can then be used to personalise insurance products and services. This can help insurers to better meet the needs of their customers and reduce churn.
Improving risk assessment: AI can be used to develop more sophisticated models of risk, which can help insurers assess the likelihood of claims better and set premiums accordingly by perceiving the customer as a segment of one. This can help to reduce losses and improve profitability.
Detecting fraud: AI can detect fraudulent claims using larger datasets with more complex patterns, which can help insurers reduce losses.
Providing insights: AI can provide insurers with insights into their business, such as trends in claims, customer behaviour, and market conditions. This information can be used to make better pricing, products, and marketing decisions.

The AI revolution is still in its early stages in Southeast Asia but can potentially disrupt the industry significantly.

In what ways has your company embraced AI technologies to improve operational efficiency or enhance business processes?

At Sunday, we have been adopting AI technology since day one. Those lie in every touchpoint process of running an insurance company, including:

Using AI for innovative new product design
Automate underwriting for better risk control
Targeted marketing
Near real-time document reading and claim assessment workflow
Fraud detection in claims

Can you share specific examples of how Artificial Intelligence has been integrated into your workforce to streamline operations or drive innovation?

An instance of AI integration in our company is evident in motor claim assessment. Traditionally, our human assessors review accident reports, analyse images of damaged vehicles, and engage in discussions with clients and repair shops to determine optimal solutions for repairing or replacing vehicle parts. However, discrepancies emerged due to variations in the skill and expertise of different assessors.

Also Read: AI must be used to enhance team members’ expertise, not to sideline them: Ravi Dodda of MoEngage

We’ve designed our AI model to perform the same task with enhanced consistency in decision-making. This AI model processes various data inputs, such as policy effectiveness, coverage details, accident reports, involved parties, garage assessments, and images of the damaged vehicle.

It utilises this data to validate key factors, including whether the car sustained damage, alignment of damage with the accident report, identification of affected parts, and assessment of severity. Subsequently, the AI model provides recommendations to both our human assessors and garages, offering insights into how to address the repair needs of our customers’ vehicles and associated costs.

Implementing this process for over two years has significantly contributed to our company’s decision-making standardisation, fostering improved relationships with garages and expediting the review and assessment of claims. As a direct outcome, claim approvals are accelerated, enabling customers to swiftly repair their vehicles and resume their journeys with minimal delay.

What challenges or concerns did you encounter when implementing AI technologies within your organisation, and how did you address them?

I think many organisations share some common challenges and concerns, such as:

Data quality: The accuracy and completeness of data utilised for training AI systems fundamentally impact their performance. Given our status as a new startup, acquiring comprehensive data might be a challenge. Hence, we must explore alternative avenues, like government-initiated open data programs, partnerships, or innovative approaches such as one-shot or few-shot learning in lieu of conventional AI model development.
Bias: AI systems can inadvertently exhibit bias, leading to unjust or prejudiced outcomes. To mitigate this, data scientists should engage with their peers and stakeholders to validate logic and address any concerns regarding training set design.
Explainability: Explaining the rationale behind AI system decisions can be complex. Establishing trust and ensuring ethical deployment necessitates gaining buy-in, sometimes through step-by-step explanations of model reasoning or comprehensive back-test analyses to demonstrate functionality and performance.
Security: AI systems can be vulnerable to security breaches, potentially resulting in data theft or system manipulation.
Cost: The development and deployment of AI systems can be financially demanding. Occasionally, AI development serves as an experiment, where patterns may not emerge, or datasets might not generalise for future predictions. Regular project assessments and decisive actions are crucial to navigate such situations effectively.

How do you ensure transparency and uphold ethical considerations in using AI technologies within your organisation to mitigate privacy concerns?

Several strategies exist to establish transparency in deploying AI technologies within an organisation, effectively addressing privacy concerns. For instance:

Develop a clear AI ethics policy: Craft a robust AI ethics policy that underscores the organisation’s commitment to transparency, accountability, and equity. Our policy addresses the following facets:

– The specific objectives of each AI/ML system
– Utilised data sources and types
– Methods of data collection and application
– Protocols for system development and testing
– Mechanisms for continuous monitoring and evaluation
– The role of the system in decision-making processes
– Strategies employed to counteract bias and discrimination

Involve stakeholders: It is important to involve stakeholders in developing and using AI systems. This includes employees, customers, regulators, and the public. Stakeholders can provide valuable input on the system’s ethical implications and help ensure that it is used responsibly.
Be transparent about the use of AI: Sunday wanted to be transparent about using AI systems as much as possible. This includes disclosing the purpose of the system, the data that will be used, and how the system will make decisions. The organisation should also provide clear information about how individuals can access and control their data.
Mitigate privacy concerns: We aim to use anonymised or pseudonymised data whenever possible and ensure the system is protected from unauthorised access.

Also Read: AI is not about job displacement but job augmentation: Nick Eayrs of Databricks

How do you ensure that AI technologies complement your workforce’s existing skills and expertise rather than replacing or displacing human workers?

We initiate by thoroughly grasping our employees’ skills and expertise. Subsequently, we outline precise objectives for the role’s evolution over time, providing a well-defined path for upskilling or reskilling. This approach ensures that integrating AI technologies harmonises with existing skills rather than supplanting them. Below are potential strategies to formulate such a plan:

Design AI technologies to be used in collaboration with humans: AI technologies should be designed to be used in collaboration with humans rather than as a replacement for humans. This means that AI technologies should augment human skills and capabilities rather than replace them.
Provide training and development opportunities for employees: Employees should be provided with training and development opportunities to help them learn how to use AI technologies effectively. This will help ensure that employees can stay relevant in the workforce and contribute to the development and use of AI technologies.
Create a culture of innovation and collaboration: A culture of innovation and collaboration will help ensure that AI technologies are used to benefit the entire workforce. This means creating an environment where employees feel comfortable sharing ideas and working together to solve problems. And in the end, we would leave one behind.
Monitor the impact of AI technologies on the workforce: It is important to monitor the impact of AI technologies on the workforce continuously. This will help identify potential problems early on and take steps to mitigate them.

How do you envision the future collaboration between humans and AI? What role do you see AI playing in augmenting human capabilities?

From my perspective, I envision the future collaboration between humans and AI as a partnership where humans and machines work together to achieve common goals.

AI will augment human capabilities by automating tasks, providing insights, and making predictions. Humans will provide AI with the context and understanding to make sound decisions.

For example, in the workplace, AI can automate tasks such as data entry, customer service, and scheduling. This can free human employees to focus on more creative and strategic work. As AI technology develops, we can expect to see even more ways in which humans and machines can work together to solve problems and create a better future.

What advice would you give to other company founders looking to leverage AI in their workforce?

In my view, AI is no longer an option, but all founders of future startups need to embrace AI, no matter what industry they are in.

Thinking it in a way as if AI is the initial substance in a chemical reaction in the next wave of industrial revolution. Companies can become irrelevant if they are missing this wave of revolution that has already started.

—

Fundraising or preparing your startup for fundraising? Build your investor network, search from 400+ SEA investors on e27, and get connected or get insights regarding fundraising. Try e27 Pro for free today.

The post AI is the initial substance in a chemical reaction in the next industrial revolution wave: Suradej Panich of Sunday appeared first on e27.