Could you tell us a little about yourself and about your studies before coming to Imperial?
Before Imperial, I completed my MEng in Mechanical Engineering at University College London. During my time at UCL, I was involved in many robotics projects, including designing and building autonomous UAVs for competitions. These projects, such as developing obstacle avoidance navigation systems and image recognition, sparked my interest in AI and ML and motivated me to pursue advanced studies in this field.
What attracted you about the MSc in AI?
The MSc in AI at Imperial stood out to me because of its perfect mix of theoretical and practical elements. It's not just about diving into theory; it's also about getting hands-on experience with software engineering for AI systems. I loved the idea of how to actually build and apply AI solutions, not just only understanding the theory behind them. Plus, being surrounded by some of the smartest and kindest people made it an easy choice.
What did you enjoy the most?
What I enjoyed the most about the MSc AI course was being part of such a vibrant and talented community. The course brings together people from diverse backgrounds, all passionate about AI, which makes for an inspiring and collaborative environment.
In terms of the taught modules, I enjoyed Natural Language Processing (NLP). The module was engaging and we were tasked with developing a classification model to detect patronising and condescending language, based on a real-life data from the SemEval competition. The hands-on nature of the coursework, combined with the opportunity to apply NLP techniques, made it a standout experience for me.
What did you find more challenging?
One of the biggest challenges was managing the amount of work and responsibilities throughout the course. With multiple projects, assignments, and deadlines happening all at the same time, it could get overwhelming pretty quickly! I found the last few weeks of the first term particularly challenging because of the exams. Juggling all those tasks definitely taught me the art of time management...and how to survive on caffeine and determination. But it's all part of the learning journey!
Could you tell us about some of your achievements on the MSc that make you proud?
I am most proud of our software engineering group project, where my group and I developed Kiraka, an adaptive speed-reading platform. The goal was to create a system that not only helps users read faster but also improves comprehension. It does this through features like real-time eye-tracking with WebGazer to adjust reading speed dynamically, AI-generated quizzes to test comprehension, and personalised analytics to track progress.
What made this project especially rewarding was the opportunity to work closely with an incredible team. Collaborating on complex software engineering problems—like integrating machine learning models, designing adaptive algorithms, and managing both front-end and back-end development—was challenging and fulfilling. It was a fantastic experience to apply the technical knowledge from our course while learning how to tackle problems collaboratively. Seeing our project came to life and watching users benefit from our work was a proud moment for all of us.
What did you do in your spare time?
In my spare time, I enjoyed staying active through sports and going to the gym. I found it to be a great way to clear my mind and stay energised. Whenever the course is not too hectic, I also worked on my personal projects and read as much as possible to keep up with the ever-changing developments in AI.
Could you tell us about the internship you took for your individual project?
I did a 4-month internship at G-Research for my individual project, and it was a phenomenal experience to witness firsthand how large-scale machine learning systems are applied with both speed and precision in the financial markets. It was gratifying to apply the knowledge I had gained during lectures to real-world projects, especially seeing how theoretical concepts like deep learning, NLP, and model optimisation translated into impactful solutions within a professional setting. Working alongside industry experts gave me a deeper appreciation for the complexity and scalability of AI systems in a production environment.
My internship focused on researching the integration of generative AI within the company's workflows through three core projects. I developed an LLM evaluation framework to streamline model selection. This can help make it easier to verify published results and to identify the most effective models for specific tasks. Next, I fine-tuned a domain-specific embedding model using synthetic data generation on company data, which achieved a significant improvement compared to the baseline and enhanced the relevance of information retrieval within G-Research's domain. Lastly, I worked on optimising RAG performance through an adaptive embedding dimension selection system. This mechanism dynamically adjusts the dimensions of the embedding model in real-time based on system load. This approach reduces median latency while maintaining high retrieval performance, thanks to reinforcement learning techniques that balance the trade-off between speed and accuracy.
G-Research is always seeking new ideas, innovative technologies, software, and approaches to help drive value in the business in the context of our fast-paced world. Being part of an environment that embraces cutting-edge advancements and continuously pushes the boundaries of what AI can achieve was inspiring.
Overall, my internship at G-Research was an eye-opening experience that allowed me to apply and expand the skills I developed during the MSc AI program. Beyond the technical work, there were plenty of social events that provided great opportunities to connect with other interns and colleagues, making the experience even more enjoyable. My mentors were incredibly supportive throughout, offering guidance and feedback that helped me grow both technically and professionally.
What have you been doing since you graduated?
After graduating, I took some time off to relax and spend quality time with my family (it was great to recharge after an intense year of studying). Since then, I've returned to a permanent position at G-Research, where I will be continuing to work on exciting projects in GenAI.
Do you have any advice for prospective students?
The MSc AI is a fantastic course, but it is super intense, so being prepared before you start is really important. Make sure you're comfortable with Python since you'll use it a lot. Also, it's worth brushing up on your maths (calculus, linear algebra, probability, and statistics). These are the foundations for pretty much everything you'll study, so being solid on them will help you settle in easier.
In general, it's never been a better time to pursue AI as a degree. It seems like every day there are exciting breakthroughs and new research coming out, opening up possibilities that weren't imaginable just a few years ago. Staying curious, reading the latest papers, and thinking about how to apply those ideas will help you get the most out of your degree and prepare for what's next in this rapidly evolving field.