Digital Assistant Ali

Source in Arabic:   ESCWA ENACT Case Study - Digital Assistant Ali, Oman

The automated assistant "Ali" is a robotic assistant trained to understand human language and respond to questions.  It is based on artificial intelligence technology, as it works through processing the natural language of humans and machine learning to provide a simulated conversation experience for human conversation. It is an advanced type of chatbot and relies on collecting historical information and using it in algorithms to create data models and identify behaviour patterns for human conversation.  The project is part the digital transformation efforts for enhancing technical readiness, quality of services, and satisfaction of beneficiaries. 

Principles supported

• Responsiveness
• Effectiveness

Technology focus

Artificial intelligence - Big data

Purpose

• Using modern technologies based on artificial intelligence for improving and raising the efficiency of services and providing a new user experience.
• Expanding channels of providing services to include various social media, to facilitate public access to services. 
• Communicating and consolidating the insurance culture in a simple language. 
• Reducing the customers need to consult with employees.
• Providing immediate responses to inquiries.

Contributors/ Partners

• Conversations and responses to inquiries in a simple language.
• Faster access to information and completion of services.
• Answering questions using the terminology of insurance to raise awareness on the insurance culture.
• Provide a new channel to access services. 

Outcomes

• Streamline complex processes in the body with advanced automated analytics.
• Increase productivity by displaying aggregated data using reports, analytical dashboards and graphs.
• Ability to monitor the implementation of strategies and plans set based on data and indicators.
• Gain a competitive advantage.

Challenges

• There is no dedicated team to manage change when implementing.
• Lack of databases in local languages.
• Lack of expertise in preparing and managing the content of the automated assistant
• Insufficient storage and processing resources in the enterprise for Artificial Intelligence applications.
• Lack of pre-application training for technical staff.
• Incompatibility of security policies and the challenges and gaps related to data sharing. 

Lessons learned

• Conformity of security policies and regulatory circulars on data sharing must be applied in the software solution. 
• When applying AI-based solutions, enough big data is needed to train the algorithm.
• The team should have skills in the application of modern technologies.
• Employees need to be acquainted with new technologies before its use.  
• Internal change management campaigns are required to reduce resistance of employees to use modern technologies. 
• Reliance on internal human resources is needed to economise on the cost and speed of implementation. 
• Quality needs to be emphasized throughout project implementation to ensure the quality of final outputs.