Work Experience
Computer Vision Engineer at AWENTIA, founder of PassaTeoria, and prior AI / teaching roles.
Experience
AWENTIA
2024 – Now
Computer Vision Engineer
- Consulted with cross-functional teams to translate business requirements into automated dataset collection systems, developing custom firmware for embedded IoT devices to capture sensor data, reducing manual data collection time by 40% and enabling scalable AI model development
- Led end-to-end development of production AI solutions from concept to deployment by studying research papers, open-source repositories, and product documentation to identify state-of-the-art approaches, creating cross-industry computer vision SDKs that integrated AI models into edge devices for client automation
- Engineered real-time image processing pipeline implementing custom pre/post-processing operators and inference tensor decoders using GStreamer, Hailo Edge AI Processors, and NVIDIA DeepStream with TensorRT and OpenVINO optimization, transforming raw model outputs into structured metadata for seamless application developer integration
- Collaborated with compiler and platform teams to diagnose and resolve complex compilation and deployment issues across multiple edge runtimes, ensuring reliable cross-platform model execution
- Deployed production-ready AI models into scalable Kubernetes environments using Triton Server, establishing CI/CD pipelines that reduced deployment time by 60% and enabled seamless integration of computer vision solutions across multiple client projects
Computer Vision EngineerEdge AI
PassaTeoria
2026 – Now
Founder & Full-Stack Engineer
- Founded and shipped PassaTeoria (passateoria.it), a cross-platform EdTech product for the Italian driving theory exam (patente A1, A, B), delivering web, Android (Google Play), and iOS from one Expo / React Native codebase
- Built a database-first content platform: 25-lesson theory course (656 sections, ~87k words), 7,139 ministerial-style true/false quizzes, and a 4,500+ term dictionary with tap-to-translate and Italian pronunciation; native content in 7 languages plus on-demand AI translation for 170+ languages
- Designed the freemium SaaS layer end-to-end: Google OAuth, Stripe Checkout on web, Google Play Billing on Android, device entitlement, rate limiting / abuse controls, and GDPR-compliant account deletion
- Owned production operations: Supabase (Postgres, Auth, RLS) for course content and learner progress, an Express API for billing and translations, Docker on GCP provisioned with Terraform, Cloudflare TLS, and a Telegram admin bot for premium grants and comment moderation
FounderFull-StackEdTech
NOVA XR
2024 – 2024
AI Integration Engineer
- Collaborated with client teams to diagnose real-time AI integration challenges, translating requirements into technical specifications and deploying low-latency inference pipelines achieving 20ms inference time for production AI vision systems within Unreal Engine
- Deployed production AI services using Docker and Kubernetes for scalable model serving and load balancing, ensuring reliable performance for client-facing applications
AI Integration Engineer
University of Bologna
2022 – 2025
Teaching Assistant
- Communicated complex AI concepts to 300+ non-technical students, translating technical specifications into accessible learning materials and collaborating with professors to align course content with industry requirements, resulting in improved student success rates
- Collaborated with Professors Simone Martini and Michael Lodi on course material preparation and assignment grading, bridging academic and practical AI applications
Teaching Assistant