Download the job spec for full details of the role.
The Wider Research Project This AI Research is required to contribute to a new commercialisation project called Amethystcare led by Prof. Vincent Wade. The project is focused on the research and development of an Generative AI (LLM based) Conversational Agent capable of providing interactions and a level of companionship for older people living at home. The user interactions will include conversational engagement with older people regarding medicine adherence, loneliness and depression detection and cognitive decline & progress tracking. The AI researcher will develop neural architectures for multimodal (speech based) conversational agent. The will involve cutting edge technique in Generative AI. More specifically it will involve fine tuning selected foundational LLMs with exemplar input/output examples and prompt engineering to develop a configurable bank of prompts that can be personalised to the requirement of each individual user.
This work will be interdisciplinary in nature, requiring consideration of theories around conversation not only from a LLM agent technology perspective, but also incorporating knowledge from established theories in the fields of psychology, social theory and appropriate health science. The research will be part of a multidisciplinary team involving healthcare professional, full staff developer, business and research experts.
Context The position will be based in the ADAPT SFI Research Centre, hosted in the School of Computer Science & Statistics, Trinity College Dublin. ADAPT is the world-leading SFI Research Centre for AI-Driven Digital Content Technology, and brings leading academics, researchers and industry partners together to deliver excellent science, engage the public, develop novel solutions for business across all sectors and enhance Ireland’s international reputation. The ADAPT Centre is hosted and coordinated in the School of Computer Science and Statistics, TCD.
The School of Computer Science and Statistics is ranked #1 in Ireland (QS Rankings) and is a proud recipient of a Bronze Athena Swan award, attained in 2021. As part of the School’s on-going actions in relation to equality, diversity and inclusion it welcomes all applications that meet the criteria below and particularly those from under-represented groups. The School offers a collegiate and supportive environment to all its staff and works to ensure that all its staff and students can perform at their best while putting in those steps that facilitate a healthy work/life balance
Main Responsibilities
Research
As part of the overall project, this AI Researcher will work on the following tasks: – Development of Framework for fine tuning and enhancing (personalising) LLM based conversational agent. The selection techniques to provide the personalisation will be consider fine tuning existing foundational LLMs, RAG, Prompt Engineering as well as more recent approaches to LLM adaptation – Development of a working prototypes to test out Gen AI conversational affordances based on use cases involving: medicine adherence, loneliness and depression detection and cognitive decline & progress tracking – Casual conversation concerning topics of the day (news items) – Report to Lead PI and Commercial Lead and collaborate closely with the multidisciplinary team.
Administrative
As a Research within a larger Research Group in ADAPT, the person will occasionally be required to engage in administrative tasks in support of the PI/Projects overall activity. This may include drafting sections of reports for funding bodies; organising a programme of suitably themed group meetings and seminars; contributing to research funding proposals; drafting of ethics applications; and other such tasks as they arise.
Person Requirements The AI Research will require a range of knowledge, skills and attributes for successful performance in the role. The successful candidate is expected to: • Have a thorough understanding of Generative AI techniques and technologies • Experience of the development of deep learning architectures for conversational AI • Familiarity with running of experiments e.g. on a high-performance compute farm • Have excellent written and oral proficiency in English • Have excellent communication and interpersonal skills • Be willing to work as part of a multidisciplinary team and learn new cross-over skills as well as transfer skills to others • Be highly organised in their work, with an ability to balance medium term and longer-term objectives in a project. • Ability to represent the group at appropriate national and international conferences
Qualifications
Candidates appointed to this role must have completed at postgraduate degree in Computer Science, preferably with extensive experience of Generative AI or Large Language Model application development in,
Knowledge & Experience (Essential & Desirable) Essential: • Understanding of and demonstrated experience working with Generative AI • Extensive knowledge and understanding of deep learning architectures and their use in speech recognition • Proficiency and experience with a range of coding languages and environments, e.g. python, PyTorch, TensorFlow, GitHub etc • Strong commitment to their own professional development • Experience with the Linux environment (especially shell/bash scripts and user management)
Desirable • Extensive knowledge and understanding of deep learning architectures • Record of using or applying open-source publishing of code • Exposure to standard tools for editing and labelling of conversations, (e.g. Adobe Suite, Elan, Anvil, Praat or others)
Benefits • Competitive salary and equity • High-end computer and peripherals • Dedicated desk in shared office space, with generous social spaces • A creative and enabling environment with impactful research • Pension and social insurance (PRSI) included • Trinity Day Nursery • Travel Pass Scheme • Bike to Work Scheme • Employee Assistance Programme • Sports Facilities • 22 days of Annual Leave • Paid Sick Leave • Training & Development • Staff Discounts locally
Application Procedure Applicants should provide the following information when applying: 1. A motivation statement outlining their interest and suitability for the position. 2. A comprehensive curriculum vitae 3. The names and contact details (e-mail) of three referees.
Deadline for applications is 31st Oct 2024.
Applications should be sent to the below people. Please include the job title in the subject line:
Note: Candidates who do not address the application requirements above will not be considered for interview.
Further Information
Informal enquiries about this post should be made to Conor McNally ([email protected]).
