Framework MPCO generates effective context-aware prompts for diverse LLMs, improving code optimization performance and scalability across real-world codebases within industrial multi-LLM environments.

Multi-LLM Artemis AI improves code performance across domains with minimal changes, delivering substantial execution-time reductions and demonstrating the potential of collaborative LLM optimisation for efficient, reliable software.

This review examines over 50 studies on LM-based code optimization, outlining key challenges, emerging trends, and future research directions to advance efficient, reliable, and practical AI-driven program performance improvement.

IEO integrates machine learning into evolutionary optimisation to speed up hyperparameter tuning, achieving substantially faster search efficiency compared to traditional methods in classification tasks.

Artemis automates data-structure exploration and transformation to enhance software performance, showing notable reductions in CPU usage, runtime, and memory across tested C++ libraries.

ARTEMIS automatically explores and tunes interchangeable data structures to boost performance, delivering consistent gains in execution time, CPU usage, and memory across a wide range of real-world Java projects.

evoML streamlines machine-learning development with automated data processing, model optimisation, deployment, and integrated multi-objective code optimisation, reducing complexity and resource demands in building and tuning custom ML models.

Artemis automates multiobjective data-structure optimisation, delivering measurable gains in execution time, CPU usage, and memory for the Guava library through cloud-based search and tuning.