EEE Alumni Stories: Gardar Orn Gardarsson

During his time at UCL, Gardar gained invaluable hands-on experience in machine learning and integrated systems, setting him up for success in his career. The MSc programme’s multidisciplinary approach prepared him for real-world challenges, from data pre-processing to developing robust machine learning pipelines. Now, at Klaki in Iceland as their Chief Innovation Officer, he applies these skills to automate complex processes in industries such as seafood processing and pharmaceuticals, creating innovative solutions that drive efficiency and sustainability.

What initially attracted you to the MSc Integrated Machine Learning Systems programme at UCL, and how did you decide it was the right step for your career?

My intention was always to complete a master’s degree in engineering, but I had been working as a control systems engineer for the processing industries in Iceland from 2015-2020 after completing my bachelor’s degree in Mechatronics Engineering. The more I became involved with process data, the stronger my conviction became that machine learning would be the key to helping these industries develop sustainably. Taught postgrad courses in the UK were soon on my radar since that would allow me to complete the degree in one year, and while browsing the UCL website one evening, I saw a promotional video for the IMLS course. It was clearly a very hands-on, practical engineering degree which suited me well, so I started looking for ways to make it work. A friend then recommended the Chevening Scholarship to me.

How did being awarded the Chevening Scholarship influence your experience at UCL and your academic journey overall?

It was certainly a financial enabler, but it also filled me with the confidence that I was meant to be at UCL and a strong intent to do well. Doing a degree amid a global pandemic was different, to say the least, but the Chevening network is strong, and we found ways to socialise and motivate each other through the challenges that came with it.

Can you share some examples of the hands-on projects or coursework during the MSc that impacted your understanding of machine learning and integrated systems?

I loved Ryan Grammenos’ IoT course, and although we couldn’t access the EEE lab to build our own IoT solution, I set up a network with DIY electronic components I had lying around at home. Building the pipeline from the field device to the cloud and applying ML analyses there tied it all together for me.

What practical skills or insights did you gain from the programme that have been most valuable in your professional roles?

Best practice in training regularisation is one skill that I think would have been hard to acquire as compacted and practically elsewhere. Yiannis Andreopoulos’ Applied Machine Learning Systems II was invaluable to that end. He had us build super-resolution models and apply them on a benchmark. We learned to visualise and monitor the training progression, which helped identify when models were not generalising, when training was oscillating, and how to handle vanishing or exploding gradients. I often see people struggling with this in my line of work, making me appreciate Yiannis’ teaching even more.

How did the multidisciplinary approach of the MSc prepare you for the challenges of integrating machine learning with control systems in industry?

In industry, you soon realise that about 80% of ML work occurs before training and inference. Successful systems require quality data, extensive pre-processing, robust data pipelines, and careful consideration of constraints. The holistic approach of the IMLS course means its students are prepared for that.

In your current role as Chief Innovation Officer at Klaki, how do you apply the knowledge and skills you developed during your time at UCL?

Klaki provides automation solutions across manufacturing industries such as seafood processing, aquaculture, pharmaceuticals, and dairy. When heavily automating a production line, we inherently remove the human element, making it crucial to automate inspection and tasks previously performed manually. Our primary focus has been on developing advanced vision systems—for instance, creating a species classification and size measurement system for whole fish aboard fishing trawlers and inline anomaly detection for quality assurance of packaged products. These efforts have earned Klaki recognition, including the Icelandic Fisheries Award for Outstanding Value Creation.

Could you describe a particular project or experience post-MSc that exemplifies how UCL prepared you for real-world industry challenges?

I think the fish species classifier is a good example. It is a stand-alone system aboard a trawler that photographs and classifies fish travelling on a conveyor at 1.5 m/s. We have about one second to perform inference and size measurement before grading the fish appropriately. The system is also connected to a cloud-based MES system, allowing operators to monitor catch information in real-time and manage processing schedules and sales accordingly.

What career paths have you seen open up for Integrated Machine Learning Systems programme graduates, based on your observations and personal experience?

I personally experienced that it opened doors to new fields. After graduation, I could have transitioned to fintech or bioinformatics—I had such offers on the table—but ultimately chose to run a business in a field I knew and loved. ML and data engineers are in high demand across industries.

What advice would you give prospective students about making the most of their time at UCL and building a successful career in machine learning and control systems?

You will have access to experienced, world-class tutors and study alongside highly motivated students. By putting in the effort and actively socialising, you will build an invaluable network of resources that will benefit you long into the future. Moreover, London itself is a bustling hub of opportunity. Take advantage of it!

Since graduating, I've witnessed how the emergence of Large Language Models (LLMs) has dramatically popularised machine learning, placing it firmly in the spotlight across industries. ML is no longer seen as niche—it is becoming integral to mainstream technology adoption. Future ML graduates will have an unprecedented advantage, leveraging powerful LLM-based tools that grant individuals capabilities comparable to an entire team of software engineers. This shift equips them to rapidly innovate, launch their own ventures, or become indispensable within established institutions. The future belongs to ML graduates—they're entering a landscape full of opportunity, influence, and potential.

Links

• Garðar Örn Garðarsson on LinkedIn
• ​​​​​​Integrated Machine Learning Systems MSc