GTA funded – Automated Software Synthesis for mHealth: Leveraging LLMs and Model Transformations

Project Highlights

1. This project leverages NLP and LLMs for automated, precise code generation, drastically reducing mHealth app development time.

2. It enables dynamic adaptation of app functionalities using real-time data, ensuring personalized and context-sensitive health recommendations.

3. Rigorous clinical validation will establish new benchmarks for AI integration in healthcare, enhancing the reliability and scalability of mHealth solutions.

Project

The convergence of Large Language Models (LLMs), Natural Language Processing (NLP), and Model-Driven Engineering (MDE) is reshaping low-code software development for mobile health (mHealth) applications. Atrial Fibrillation (AF) requires continuous and personalised management, yet existing digital health solutions often fail to dynamically adapt to individual patient needs and changing environmental conditions. This project will enhance the P-STEP app, a platform that promotes personalised physical activity, by integrating agent-based frameworks, automated code generation, and adaptive system modelling.

The research will focus on: (1) Automated Code Generation, using LLMs to translate clinician and patient inputs into structured, model-driven components; (2) Dynamic System Modelling, leveraging MDE to create context-aware mobile applications that adapt in real time to user data and environmental conditions; and (3) Clinical Validation, ensuring the system meets medical standards and enhances patient outcomes.

An NLP-powered agent-based framework will interpret human-readable requirements, transforming them into domain-specific models that define personalised interventions. These models will be converted into deployable mobile applications, integrating real-time health data from wearable sensors and environmental monitoring systems. The system will undergo simulated validation before progressing to clinical evaluation with AF patients.

The P-STEP platform provides a test bed for AI-powered MDE automation in healthcare. This project will explore low-code software development, model-driven transformations, and adaptive architectures, using YAMTL for efficient, incremental model transformations and EMF-Syncer to maintain dynamic consistency between runtime data models and system behaviour. By validating these techniques in a real-world setting, this research will contribute to high-impact outputs and strengthen future funding applications in AI-driven software engineering and digital health automation.

A PhD candidate will gain expertise in AI-driven software engineering, agent-based frameworks, and LLM-enhanced automation for mobile health applications, preparing them for careers in intelligent automation, AI-driven software engineering, and digital health technology innovation.

Project enquiries to supervisor: Dr Artur Boronat artur.boronat@leicester.ac.uk

General enquiries CMSpgr@le.ac.uk

Please carefully read the information on our web page before applying

How to Apply https://le.ac.uk/study/research-degrees/funded-opportunities/computer-science-gta

There are 3 GTA studentships available. You can only apply for one project