Staff spotlight - Dr Swarna Chetty
Posted on Thursday 2 April 2026
1. Can you tell us about your research background and specialities?
My research background is in telecommunications and applied AI, focusing on AI-driven optimisation for 5G and emerging 6G networks, including areas such as Open RAN, network slicing, and autonomous network management.
More recently, my research has shifted toward AI safety and assurance, particularly exploring how AI models can be safely deployed in real-world systems. My current work focuses on uncertainty-aware AI, with a growing interest in subjective logic to quantify belief, uncertainty, and confidence in AI decisions.
Overall, my work sits at the intersection of telecoms, applied AI, and AI safety, aiming to ensure AI-enabled systems are trustworthy, reliable, and safe for real-world deployment.
2. What inspired you to come and work at the CfAA?
While working on AI-driven telecom systems, I started thinking more about what happens when these AI models move from research into real-world deployment. It became clear that building intelligent systems is only part of the challenge, ensuring they are safe, reliable, and trustworthy is equally important.
The CfAA really appealed to me because it focuses exactly on these questions. It also offered a great opportunity to learn from an interdisciplinary team and apply my background in telecoms and applied AI to broader AI safety and assurance challenges in real-world autonomous systems.
3. What does your role here entail and what are you currently working on?
As a Research & Innovation Associate in Safe Autonomous Systems at the CfAA, my role focuses on investigating how we can build confidence that AI models are safe for deployment in real-world autonomous systems.
Currently, I am conducting research on subjective logic and uncertainty-aware AI. Traditional machine learning metrics such as accuracy and recall measure model performance, but they do not fully address how confident we can be in those models, particularly in safety-critical environments.
Subjective logic provides a framework to explicitly represent belief, disbelief, and uncertainty in AI predictions. This enables a more comprehensive understanding of model reliability and supports AI assurance, helping determine when AI systems are sufficiently confident and safe for deployment, and when additional safeguards or human oversight may be required.
4. What have been your career highlights so far?
Apart from my research, I’ve also been involved in innovation and entrepreneurial activities. I served as the Entrepreneurial Lead for RIC Connect through the UKRI ICURe Explore programme, where I conducted extensive market discovery, engaging with telecom operators, vendors, and regulators to understand real-world needs and validate the commercial potential of the technology. Through this programme, I also received training for initiating a spin-off and was recommended to progress the technology towards a potential spin-out, which was a valuable experience in translating research into real-world impact.
I have also been co-author of multiple award-winning papers, including Best Paper awards at international conferences. In addition, I was recognised among the “100 Brilliant and Inspiring Women in 6G” in 2025 and 2026, highlighting my contributions to AI-native 6G and next-generation communication systems.
5. Finally, where can we find you outside the CfAA?
When I’m outside of work, you can usually find me in one of two places: wandering through the middle of nowhere in search of a hidden hiking trail (and the cozy pub that inevitably follows) or in my kitchen, attempting to recreate a dish I tasted once and now, perhaps overconfidently, believe I’ve mastered. (Spoiler: I haven’t, but the effort is always high-energy!) I’m a massive foodie, and my latest obsession is Mexican cuisine. This has resulted in some truly excellent salsa and tacos, along with a few "experimental" batches that were questionable at best, though I bravely ate them anyway in the name of science. Speaking of science, my baking sessions often leave the kitchen looking like a laboratory mishap, but the results are usually delicious. If anyone is looking for an excuse to eat Guinness cake, please ping me, I am professionally trained in finding reasons to bake one! To balance out the chaos of the kitchen, I love to get creative with crocheting and painting. I’m a fan of starting wildly ambitious projects and proudly finishing them...eventually. When the sun goes down, I’m all about jazz, a good glass of wine, and my annual tradition of learning something new. This year’s challenge? Latin dance. So far, my progress consists of 10% rhythm and 90% mild confusion, stepping on toes, and counting "one-two-three" under my breath with intense focus. I’ve also been falling down a deep mystery rabbit hole lately. Between Agatha Christie novels and binge-watching Miss Marple or Poirot on Netflix, I spend a lot of time confidently “solving” the crime halfway through only to be proven wrong most of the time. When I need a bit of fresh air or at least the indoor equivalent, you’ll find me tending to my ever-growing collection of houseplants. I’ve been experimenting with plant propagation and grafting lately, though my success rate is best described as "optimistic." Some of my green roommates thrive, while others choose to give up quite dramatically. But hey, the living room is slowly becoming a jungle, and I’m learning something new with every leaf!