Lead ML Engineer @ Deloitte | Physical AI, Agentic AI & Real-Time Computer Vision | 7x NVIDIA Certified | CS Graduate, FIB-UPC
I lead a small engineering team at Deloitte building production Physical AI, Real-Time Computer Vision, and Agentic AI / RAG systems — from factory-line anomaly detection and camera-tracked digital twins on NVIDIA Omniverse & Isaac Sim to multi-cloud inference and data platforms across AWS, Azure, and GCP. Computer Science graduate from FIB, Universitat Politècnica de Catalunya (UPC).
- FleetMind — NVIDIA EMEA Physical AI Hackathon (1st Place) — Team win across EMEA: AI-driven warehouse orchestration PoC combining NVIDIA NIM, cuOpt, Isaac Sim, and conflict-based search.
- polcasacubertagil.com — Astro 7 + React 19 (TypeScript) portfolio with 20 interactive ML, CV, and algorithm demos, EN/ES/CA i18n, Sentry, Playwright/Vitest, and a CI-driven LaTeX CV pipeline.
- TFG — Colonoscopy Polyp Detection — FIB-UPC Bachelor's Thesis (9.2/10): Faster R-CNN, RetinaNet & SSD augmented with CycleGAN, SPADE, and Stable Diffusion + LoRA.
- bitsXlaMarató 2022 Winner — Team win (cardiovascular health challenge): Mask R-CNN abdominal aortic aneurysm detector on ultrasound video with 3D aorta reconstruction.
- dotfiles — GNU Stow Linux/macOS setup (~2,000 commits, Shell & Python tests, Gitleaks & CI).
- Open-source PRs: Two open pull requests (#2496, #2498) to
vercel-labs/skillsaddingdisable-model-invocationfrontmatter & CLI support.
7x NVIDIA Certified (Professional: Agentic AI, Gen AI LLMs, OpenUSD Development, Accelerated Data Science; Associate: Generative AI Multimodal, Generative AI LLMs, AI Infrastructure & Operations) · Google Cloud Generative AI Leader · Microsoft Azure Fundamentals (AZ-900)



