Cambridge, United Kingdom
Ioannis Politis
Staff UX Researcher at Arm · AI and developer experience · PhD in HCI
About
I am a user experience researcher with a PhD in human-computer interaction and fifteen years of work on how people use complex tools: cars, medical devices, programming languages and now AI infrastructure. What I keep asking is what a person can perceive, remember and decide at the moment a system requires it.
At Arm I lead research on developer workflows for large-scale AI: how engineers find, judge and deploy machine learning models. Before that, six and a half years at MathWorks on the user experience of writing code, two years at the University of Cambridge on trust and handover in autonomous vehicles, and a PhD in Glasgow on warnings a driver could hear, see and feel.
I work across quantitative and qualitative methods, and the studies I care most about put the two together: a number showing how widespread something is, next to a session showing why it happens.
I grew up in Athens, studied in Piraeus and Eindhoven, and have been in Cambridge for a decade. Away from research I take photographs and play music.
At a glance
- Role
- Staff UX Researcher, Arm
- Focus
- AI and developer experience: model discovery, evaluation and deployment workflows
- Based in
- Cambridge, United Kingdom
- Methods
- Survey design and benchmarking, statistical analysis, usability testing, interviews, diary studies, mixed methods
- Domains
- AI platforms and developer tools, programming languages, autonomous vehicles, medical devices
- Tools
- AI agents as my main working layer, over Python, SPSS, MATLAB and SQL
- Education
- PhD in HCI (Glasgow), PDEng User-System Interaction (TU Eindhoven), MSc (Piraeus), BSc (Athens)
- Research
- 40+ peer-reviewed publications and book chapters, listed on Google Scholar (opens in a new tab)
- Languages
- English, Greek, German (working)
- Contact
- idpolitis@outlook.com and LinkedIn (opens in a new tab)
How I work
Six things that have held across fifteen years, four countries and every domain I have worked in.
-
Interfaces that are not screens
Sound, vibration, code, command lines. My PhD designed warnings a driver could feel. At MathWorks the object of study was a function signature and the error it returns. At Arm it is a deployment that either confirms it worked or leaves you guessing. Most usability methods assume there is something to look at, and much of my work has had nothing.
-
Measurement with statistics behind it
Controlled experiments and large-scale surveys, analysed properly: ANOVA, regression and non-parametric tests across sub-groups, in Python, SPSS and MATLAB. Five simulator experiments in the PhD, global surveys across five product teams at MathWorks, multi-wave practitioner surveys at Arm. Forty peer-reviewed papers means the analysis has been checked by strangers.
-
Building the instrument
Some questions have no measure yet, so I have built the thing that measures. A tactile warning vocabulary. An acceptance model for self-driving cars that other researchers now use. A set of data-driven user profiles that gave AI teams one vocabulary for their users.
-
Where mistakes are expensive
Regulated medical imaging, cars at speed, AI infrastructure. In these settings you have to be able to show that a design is safe to use, not simply believe it. Usability testing in hospitals across three countries taught me what that costs and how to plan for it.
-
The people usually left out
Inclusive user groups for autonomous vehicles, autistic drivers, older users, blind navigation, children in cerebral palsy therapy. This runs through my work from 2010 to now. The people who stand to gain most from a new system are usually the last to be recruited into a study of it.
-
Research a whole team can do
Standards, templates and review so that designers, engineers and product managers run sound tactical research themselves. I have set this up inside two engineering-led organisations. It multiplies how much a company can learn and frees a researcher for the questions that need one.
Selected work
All work →
Experience
- 2025 to nowStaff UX Researcher, Arm
Research strategy and execution across AI and developer tooling product areas.
- 2018 – 2025Senior UX Researcher, MathWorks
Quantitative and mixed-method research across five or more product teams building AI and deep learning tools.
- 2016 – 2018Research Associate, University of Cambridge
Human-computer interaction for autonomous vehicles at the Engineering Design Centre.
- 2012 – 2016Doctoral researcher, University of Glasgow
Multimodal audio, visual and tactile driver displays, funded by Freescale Semiconductor.
- 2011 – 2012Interaction Designer and Usability Engineer, Philips Healthcare
Interaction design, UI standards and usability testing for medical imaging systems.
- 2009 – 2011PDEng trainee, User-System Interaction, Eindhoven University of Technology
Inclusive, playful and accessible interfaces, from assistive devices to therapy games.
Education
- 2012 – 2016PhD, Human-Computer Interaction, University of Glasgow
- 2009 – 2011PDEng, User-System Interaction, Eindhoven University of Technology
- 2008 – 2009MSc, Advanced Information Systems, University of Piraeus
- 2001 – 2007BSc, Informatics and Telecommunications, University of Athens