Empowering Research with AI: Highlights from the First 2025 Techkie Workshop

Our first Techkie workshop of the year launched with an exciting collaboration between Women in AI UK, IEEE Women in Engineering UK and Ireland, and the IEEE Intelligent Transportation Systems Society. Held in the Peter Landin Building at Queen Mary University of London, this was also the first major activity led by the QMUL IEEE Student Branch, under the guidance of Dr. Mona Jaber. The workshop brought together over 13 enthusiastic participants from institutions including Nottingham University, the University of West London, the University of Exeter as well as industrial participants, all eager to explore how AI can reshape their research.

Led by the inspiring Yasmin Fathy, the workshop addressed real-world challenges in the transport sector—particularly the issue of imbalanced datasets. We explored the power of Generative Adversarial Networks (GANs) and Conditional GANs (CGANs) to generate synthetic data that balances underrepresented classes. GANs, which are especially effective in research scenarios with limited data, pit a generator and discriminator against each other to produce remarkably realistic outputs. This technique helps create richer, fairer datasets that can significantly enhance model training for critical areas like road safety.

In the second half, we rolled up our sleeves and dove into Large Language Models (LLMs)—specifically LLaMA-based RAG pipelines. Yasmin walked us through web scraping, text chunking, embedding creation, and context retrieval for intelligent Q&A generation. A huge thanks to our student team: Chair Kaiwei Wang, Vice Chair Tingting Yang, and Secretary Jose Ricardo Treviño Rodriguez, who also served on-site as a teaching assistant, ensuring everything ran smoothly. We’re grateful to AI4BetterLife UK and Queen Mary University of London for their support—and we can’t wait for the next Techkie session!