ML ENGINEER · RESEARCHER
CARLO ZAMUNER
APPLIED AI · RESEARCH SYSTEMS
MACHINE LEARNING ENGINEER AND RESEARCHER WITH FOCUS ON FOUNDATION MODELS AND COMPUTER VISION. BUILDING RELIABLE AI SYSTEMS FROM RESEARCH PROTOTYPES TO PRODUCTION.
PROJECTS
Playable case studies
SwimMind Control Room
Predictive sets, attendance automation, and Bluetooth timing integrations for club coaches. AI prompt pipeline, using langchain and logfire.
Home Repair Assistant
Agent-to-Agent protocol orchestrating five autonomous agents across diagnosis, DIY, scheduling, and feedback.
3D Pose Estimation & MoCap Align
Triangulated multiview keypoints to recover basketball poses and validated against MoCap at 18 mm MPJPE.
CLIP-Muon Few-Shot Adaptation
Fine-tuned CLIP with Muon + LoRA adapters to hit state-of-the-art few-shot performance.
Sionna × Unreal RF Lab
Interactive RF experimentation toolkit coupling UE5 visuals with differentiable comms blocks.
ABOUT
Who I am
Currently leading ML research at SwimMind and collaborating with labs on LLM evaluation, multimodal sensing, and GPU optimization workflows.
CURRENT FOCUS
SwimMind AI research & copilots
STACK
PyTorch · CUDA · LangChain
BASE
Trento · Remote research labs
EXPERIENCE
Where I ship impact
SwimMind Platform
CURRENTFounder · Full-Stack AI Engineer
Aug 2025 – Present
Building an AI operations cockpit for competitive swimming clubs with realtime analytics, Gemini copilots, and Google Cloud automations.
Industrial AI Challenge · HIT
AI Challenge Participant
Sep 2025 – Dec 2025
Engineered energy-demand forecasting models, biogas + solar simulations, and RL optimization that cut emissions by 15%.
Nvidia Sionna Lab · University of Trento
GPU-Accelerated AI Research Intern
Apr 2024 – Sep 2024
Linked Unreal Engine with Nvidia Sionna and optimized CUDA pipelines for RF experimentation.
CONTACT
Ready to collaborate?
I collaborate with founders, studios, and labs to ship AI products end-to-end. Open to opportunities in ML research, computer vision, and production AI systems. Reach out and let's build.