Air India to help US, European airlines create AI chatbots
Air India is already in talks with a number of leading US and European airlines. These companies are looking for help in creating generative AI chatbots, just like Air India's own AI.g. The Indian airline's sophisticated chatbot has been extremely efficient, managing nearly 97% of all customer queries on its own. This level of automation greatly minimizes the requirement for human contact center agents.
AI.g's impressive performance attracts global interest
AI.g, which was launched in May 2023, has handled approximately four million queries since its launch. The chatbot receives some 30,000 queries per day on average, and deals with almost all of them without the need for live agent support. This remarkable performance has caught the attention of international airlines, which are now in talks with Air India for similar projects.
Air India official shares insights
Satya Ramaswamy, Air India's Chief Digital & Technology Officer, has said that the company is focusing on innovation to make AI.g stand out. Speaking at a media briefing, he revealed that several US and European airlines have approached them for help in developing similar AI chatbots. Ramaswamy emphasized the uniqueness of their chatbot technology, noting that despite sharing knowledge about it, no one has been able to replicate their success.
Air India files for 1st patent
Along with its chatbot technology, Air India has also applied for its first-ever patent. The application is for an upcoming 'one click booking' feature, which is expected to be offered on the airline's website and mobile app by December. This move marks a significant milestone in the company's history, as it undergoes a transformation phase under Tata Group's ownership.
Ensuring chatbot safety and unbiased responses
Ramaswamy also addressed concerns regarding potential racial bias in the chatbot's responses. He assured that measures are being taken to ensure the chatbot is "safe to use, does not have bias and limits (responses) to the queries that have been asked." This includes using Large Language Models (LLMs) for training, but also employing innovative strategies to prevent any form of discrimination or bias in its operations.
