Work with me
I supervise MSc theses and semester projects at ETH Zürich, and collaborate widely. If you want your models to run on real hardware and your name on real papers, read on.
Why join
Top-venue publishingRecent student work landed at NeurIPS, EMBC, and IEEE journals.
Real hardwareGAP9, RISC-V platforms, BioGAP, and wearable prototypes, not just simulations.
Close mentorshipWeekly 1:1s, clear milestones, and career guidance beyond the thesis.
Strong networkCollaborations across ETH, EPFL, University of Bologna, and industry.
Open topics
MSc Thesis
Tiny Recursive Models for Time-Series
Adapt TRMs to EEG and UCR benchmarks; study how deep supervision and adaptive halting trade accuracy against compute.
MSc Thesis
Quantized TRMs for Edge Deployment
INT8/INT4 TRMs on GAP9 and Cortex-M: accuracy and energy trade-offs, on-device adaptive halting.
+
Propose your own
Foundation models, TRMs, NAS, or TinyML deployment: pitch an idea at the intersection and we will scope it together.
How it works
Reach outEmail me with your CV, transcript, and a few lines on what excites you.
ChatA 30-minute call to find the overlap between your interests and open directions.
ScopeWe define the project together: research question, milestones, hardware.
StartWeekly supervision, compute access, and a paper-shaped goal from day one.