Andrea Simonelli

  1. Now NVIDIA Senior Solution Architect Zurich, Switzerland
  2. 2022 – 2026 Meta Senior Research Engineer Zurich, Switzerland
  3. 2018 – 2022 PhD Computer Science University of Trento

I work on applied AI, at the intersection of research, engineering and product. I'm currently a Senior Solution Architect at NVIDIA, supporting startups and large companies across EMEA on large-scale data processing and training, vision-language models (VLMs) and action models. Before that I spent four years at Meta, most recently as a senior research engineer, working on 3D scene understanding and fine-tuning vision-language models. During my PhD I interned at the startup Mapillary, working on 3D perception at scale, and worked part-time as a solution architect at FBK on projects with companies including Generali.

Andrea Simonelli

Experience

  1. NVIDIA2026 – now

    Senior Solution Architect, Physical AI

    I support startups and large companies across EMEA on large-scale data processing and training, vision-language models (VLMs) and action models, through proofs of concept and technical support. I also use agentic AI to optimize and automate processes, and run technical training and hands-on sessions at conferences and events.

  2. Meta2022 – 2026

    Senior Research Engineer Promoted to Senior in 2024

    3D scene understanding with transformers and vision-language models. Fine-tuned open-source VLMs for 2D and 3D perception and visual question answering, and evaluated them on VQA and detection benchmarks (research work). Led the R&D of an interactive 3D segmentation method for Meta Hyperscape, optimized for real-time inference on limited hardware. First author of Easy3D (ICCV 2025, oral).

  3. Meta2021

    Research Intern

    Led a project bridging 3D detection and 3D reconstruction, which resulted in an internal automatic 3D annotation tool. Co-author of AutoRF (CVPR 2022).

  4. FBK2018 – 2021

    Solution Architect, 50% during the PhD

    Worked with companies including Generali (Milan) on proofs of concept and production-level algorithms, for example for automatic insurance claims and an automatic retail system. Led the development of several machine-learning systems and deployed them on low-cost hardware such as NVIDIA Jetson.

  5. Mapillary2019

    Research Intern

    Developed a product for automatic scene understanding and mapping based on 3D perception, trained and deployed on Mapillary's crowdsourced data (3B+ images). Led a research project on 3D detection, published at ECCV, which won the first nuScenes challenge (camera-only).

  6. University of Trento2018 – 2022

    PhD in Computer Science

    "Bright Minds" industrial grant, co-funded by Mapillary Research and FBK. Obtained with honors. Teaching assistant at the university during the PhD.

Get in touch

The best way to reach me is email.

simonelli.andrea23@gmail.com LinkedIn Google Scholar