Introducing the Quris AI chatbot

Artificial Intelligence’s (AI) rapid expansion throughout the 21st century prompted many enquiries over its role in transforming the healthcare system. AI is already being used across the healthcare industry, from detecting diseases and identifying early treatments to providing clinical support and guidance to healthcare workers.

With the role of artificial intelligence continuing to evolve and transform healthcare, we decided to design our own Chatbot service - Quris - into our Quris Platform. Unlike many other applications of AI, ours is one that can be trained for precise medical use, with capabilities specifically tailored to healthcare needs.

With the help of Microsoft services, our Chatbot is driven by the power of Co-pilot and automation Agents within Microsoft Azure AI cloud services to maximise efficiency and reliability in data outcomes, i.e. the results of interactions with the AI Chatbot.

With the power of Microsoft Co-Pilot, which combines the power of large language models (LLMs) embedded in the most commonly used Microsoft Office 365 apps, productivity and creativity within the workforce can be maximised.

LLMs utilise broad datasets, but the real value for businesses comes in connecting these models to your specific business data, which is exactly what Co-pilot is designed to do. By providing the most up-to-date and relevant information to its user and by harnessing the massive reservoir of data and insights lying largely untapped, Co-pilot provides people more agency and makes technology more accessible to support staff within healthcare settings in their daily roles. 

Despite the productivity gained from applications such as Co-pilot, issues surrounding complex and time-consuming workflows continue to limit the full potential of its application across businesses.  This is a challenge we sought to overcome and that is where Azure’s AI Agent’s come in. 

Azure AI Agent Service uses autonomous AI assistants to automate many business processes – improving results and empowering teams.

Agents can manage conversation state and connect with internal and external knowledge sources to ensure the right context to complete a process, shifting traditional generative AI responses based on generic data to contextualise specific answers directly related to your enterprise. 

To help optimise business processes and enhance productivity, there are three levels of Agents :

Level 1: Retrieval ~ Information is retrieved from grounding data, reasoned and summarised to answer user questions.

Level 2 : Task ~ Actions are taken when asked, this can automate workflows and replace repetitive tasks for users

Level 3 : Autonomous ~   Agent operates independently to dynamically plan and orchestrate other agents to learn and escalate the user’s needs.

The three levels of agents enable a business to tailor their AI assistant to the job at hand on the basis of the complexity and capabilities required.

Our chatbot service is already rolled out within the Manchester University NHS Foundation Trust and is being used to support workforce wellbeing and reduce administrative demands on busy departmental resources.

Looking forward, we are actively seeking to develop our chatbot features by incorporating transactional services, allowing staff answers to questions specifically related to their individual needs.

The latest development of our chatbot capabilities will be deployed within the context of clinical decision support to deliver capabilities such as review, updating and improvement of clinical guidelines thereby freeing up staff time to focus on front-line care delivery.