Tag AI assistant

SEOUL, August 26, 2024 – SK Telecom, announced a complete overhaul of its innovative artificial intelligence (AI) service, ‘A.’ (A-DoT). The revamp aims to transform the A. app into an AI personal assistant focused on enhancing daily convenience for customers.

The most significant feature of this revamp is the enhancement of natural conversation experience based on a large language model (LLM) and the strengthening of daily management functions through multi-agents. The entire user experience (UX) has been overhauled to enable flexible conversations with an LLM-based agent.

A. has integrated all essential functions into the new ‘Daily’ feature, including calendar, tasks, schedules, routines and sleep. This enables comprehensive management of all aspects of a user’s daily routines.

Users can save and manage appointments, meetings, to-dos and more through voice commands, reducing the need for conventional, tedious typing. Moreover, the new service greatly enhances the daily AI personal assistant experience by providing customized suggestions tailored to the user’s personal situation, preferences and other factors, such as alerting the user to weather or traffic conditions before their appointments.

In the updated A. app, users can access an array of seven multi-LLM agents. These include Perplexity, A.X, ChatGPT 3.5 turbo, ChatGPT 4o, Claude Haiku, Claude Opus and Claude Sonnet. Each agent offers unique conversational AI capabilities tailored to the user’s specific needs. Users can select an LLM engine that aligns with objectives, enabling them to compare responses from various models for the same query. These features will be provided free of charge for the time being.

In addition, the app provides specialized agent services across various fields, including music, media and stocks, ensuring a comprehensive and immersive user experience.

“The A. overhaul ensures a more natural and convenient conversational experience as if the users are interacting with a real personal assistant,” said Kim Yonghun, Vice President and Head of the AI Service Business Division at SKT. “We will continue to innovate to provide AI services that save our customers’ precious time and enhance convenience and quality of life.”

Dora, an AI-powered automated speech system, can call patients, ask them questions, comprehend their responses, and recognise with precision which ones call for a clinical review. Researchers from Imperial College Healthcare NHS Trust, Oxford University Hospitals (OUH) NHS Foundation Trust, and Newcastle University conducted the study, which was published in the open access journal eClinicalMedicine of The Lancet. The study included more than 200 patients who had recently undergone cataract surgery, the most common procedure done by the NHS. Three weeks following the procedure, Dora called to ask them to perform a follow-up evaluation, which was observed by an ophthalmologist specifically for the study. Researchers evaluated the independent judgements on the clinical relevance of five symptoms and whether the patient needed more evaluation made by Dora and the supervising ophthalmologist. Dora’s choices throughout the trial closely matched those made by the physicians.

The study’s co-principal investigator, Newcastle University Professor of Digital Health and Clinical AI, Edward Meinert, led the acceptability assessment. He said: “The way in which patients responded to the AI system suggests that it may be expanded and used in other clinical paths, potentially saving the NHS time and money. The study assessed the cost impact of Dora in comparison to traditional treatment, as well as its usability and acceptance to patients. The AI follow-up was accepted by the majority of participants, however some expressed worry about the absence of a “human element” in situations with problems.”

Co-Principal Investigator at OUH and honorary consultant ophthalmologist Kanmin Xue stated: “The Dora system is able to call patients autonomously by telephone and conduct a conversation that gathers the key information needed to determine whether a patient is recovering well or requires further medical assessment.” Those that required further clinical input were successfully identified and given a higher priority. Thousands of patients at OUH have had their post-cataract surgery assessments conducted using an updated version of Dora, which was implemented following the trial. We are now using it for pre-operative evaluations as well, in which Dora asks the patient a series of screening questions, one of which is whether or not they want surgery for their cataract. The information is recorded in our electronic health records, which expedites the in-person pre-assessment guided by a nurse. Compared to traditional care, using this telemedicine follow-up examination was projected to have saved each patient about £35 in costs. “

Health Innovation Oxford and Thames Valley provided assistance for the development of Dora by the Oxford spin-out business Ufonia. These organisations conducted a feasibility study, examined physician comments, and assessed the financial benefits of integrating Dora into the post-operative patient route. CEO of Ufonia, Dr Nick de Pennington, stated: “This research demonstrates unequivocally Dora’s high accuracy in identifying patients who require human clinical input.” We are currently extending Dora’s “Artificial Clinical Intelligence” throughout the National Health Service’s and other countries’ eye care pathways. In terms of cost- effectiveness, acceptability, and safety, the system provides a lot of advantages. In addition, it  helps hospitals lower their carbon impact, and most significantly, it makes healthcare more accessible to patients.”

The National Institute for Health and Care Research (NIHR) Artificial Intelligence in Health and Care Award is providing funding for this study.