Franco Caspe professional portrait

I am a PhD candidate at QMUL's AIM CDT, a musician, maker, and analog photographer.

I am interested in expressive, and nuanced interactive music technology that is context-aware, and feel natural to use. At QMUL, I am designing novel sound-based control modalities for musical instruments with AI .

I also have an International M.Sc. Degree in Image Processing and Computer Vision; my thesis was about Real-Time DNN serving infrastructure, done at MPI-SWS.

Previously, I studied in Argentina where I received my diploma in Electronic Engineering while working in R&D on Defense and Communications. During that time I also studied music and played with several bands in my city, playing guitar, keyboards and singing.

I have interned in Yamaha Music, Max Plank Institute for Software Systems, and with the IAESTE programme.

Latest News

Date Update
11th Nov 2025 Presenting a new talk at ADC 2025, Engineering Practices Break Music Interaction, where I share the main takeaways of my PhD journey designing AI systems for low-latency musical interaction.
23rd Oct 2025 Presenting our paper in Low-Latency Neural Synthesizers at the AES Show in Long Beach, California.
13th Oct 2025 In-person demoing of Minifusion: Live Instrument Transformation at WASPAA in Lake Tahoe, California, USA.
10th Oct 2025 I am giving a talk on Neural Audio Synthesis for Low-Latency Interaction at the CIRMMT's GAMMa Workshop
25th Sep 2025 Streaming online our tutorial on Real-time Neural Audio Inference and Design with Jatin Chowdhury!
8th Sep 2025 Giving a Tutorial on Real-time Neural Audio Inference and Design with Jatin Chowdhury at the first AES conference in AI and ML for Audio (AIMLA) in London.
24th Jun 2025 I am presenting my paper and demo on Waveform Autoencoding for Instrumental Interaction at NIME 25 in Australia
12–14th Jun 2025 Showcasing the BRAVE Plugin at Sonar 2025 in Barcelona!
May 2025 Excited to go for a month placement in Iceland, at the Intelligent Instruments Lab! Check out this Collaboration we did with Nicola Privato combining the BRAVE model and his great Stacco Instrument.
11th Apr 2025 Our paper Designing Neural Synthesizers for Low-Latency Interaction has been published in the May 2025 Issue of JAES.