Singapore Airlines integrates OpenAI models for customer service and internal operations

Overview of AI Integration at Singapore Airlines

Singapore Airlines has integrated OpenAI’s models across its operations to enhance customer-facing services, streamline internal workflows, and improve employee productivity. By utilizing the OpenAI API alongside a governed internal platform named Jarvis, the airline has transitioned from rigid, scripted systems to flexible, AI-powered conversational experiences and development tools.

Conversational Customer Support

The airline replaced its traditional dialog-flow chatbot, Kris, with an LLM-powered conversational interface. This transition removed the need for manually mapping every possible conversation path, allowing the system to interpret natural language, ground responses in a knowledge base, and manage multi-topic interactions.

Key performance metrics for this deployment include:

  • 2x increase in customer satisfaction (CSAT) scores.
  • ~10% reduction in average handling time for customer service representatives, supported by AI-powered call summarization.

AI-Driven Travel Discovery

Singapore Airlines launched a "Flight Recommender" tool that allows customers to describe travel preferences in natural language rather than using traditional search filters. The tool utilizes an orchestration layer to connect to fare and availability data, interpreting intent and returning personalized recommendations. The airline reports that usage of this tool has grown nearly 3x over the past year. Future updates aim to transition this into a multi-turn conversational experience where users can refine travel plans without restarting their search.

Software Development Lifecycle Productivity

Beyond customer-facing applications, the airline has integrated generative AI into its software development lifecycle (SDLC). Teams are using these models to assist with requirements refinement, stakeholder analysis, test-case creation, and architecture reviews.

According to the airline, this integration has led to:

  • Up to 20% productivity gains in selected development activities.
  • A shift in team roles from manual artifact creation to directing, reviewing, and validating AI-generated outputs.
  • Improved requirements quality and test coverage through the identification of edge cases before development begins.

Internal AI Platform: Jarvis

To standardize AI access across the organization, Singapore Airlines developed Jarvis, a governed internal platform. Jarvis allows employees to interact with various file types—including PDFs, Word, Excel, and PowerPoint—in a secure environment. The platform supports the creation of custom workflows and lightweight knowledge bases. Currently, over 60% of Singapore Airlines employees use Jarvis on a monthly basis.

Sources