Hey there 👋

I'm Federico Betti

Researcher, AI Engineer and Startup Enthusiast

I'm a Computer Engineer with a focus on AI research and development, currently pursuing my PhD at the University of Trento while serving as Senior AI Engineer at Typewise - ETH Zurich. My work spans natural language processing, generative AI, and efficient ML implementations.

Let's connect 📄 Download CV

Current Position

See others →

Senior AI Engineer at Typewise - ETH Zurich

As a Senior AI Engineer at Typewise (Y Combinator alumni), I work at the intersection of industry and academia through our collaboration with Prof. Ryan Cotterell's NLP lab at ETH Zurich. I lead the development of advanced text prediction and grammar correction systems, strategically balancing computational efficiency with model performance to create production-ready NLP solutions. My responsibilities include optimizing large language model integrations, architecting scalable AWS infrastructure, and ensuring robust deployment of our language technologies. Recently, I've been focusing on developing AI agents for customer care applications, leveraging cutting-edge libraries and frameworks to create responsive, context-aware solutions. Beyond technical leadership, I actively participate in strategic discussions with the founding team on product development, business growth, and organizational scaling, including talent acquisition and team development.

NLP Text Prediction Grammar Correction Large Language Models AWS Performance Optimization Product Development YC Startup AI Agents Customer Care

Recent Research

What Changed? Detecting and Evaluating Instruction-Guided Image Edits with Multimodal Large Language Models

This paper introduces DICE (DIfference Coherence Estimator), a model designed to detect localized differences between original and edited images and assess their relevance to modification requests using Multimodal Large Language Models.

ICCV 2025 | Published: May 2025

Authors: L Baraldi, D Bucciarelli, F Betti, M Cornia, L Baraldi, N Sebe, R Cucchiara

View PDF

Computer Vision Image Editing Multimodal LLM Evaluation Metrics Human Judgment DICE Instruction-Guided Generative AI Self-Supervision Inpainting

Optimizing Resource Consumption in Diffusion Models through Hallucination Early Detection

We introduce HEaD (Hallucination Early Detection), a novel paradigm that swiftly detects incorrect generations in diffusion models during early stages, saving computational resources and accelerating the generation process for complex multi-object images.

Proceedings of the European Conference on Computer Vision Workshops. 2024. | Published: January 2024

Authors: F Betti, L Baraldi, L Baraldi, R Cucchiara, N Sebe

View PDF

Computer Vision Diffusion Models Resource Optimization Computational Efficiency Hallucination Detection Multi-object Generation Cross-attention Maps Energy Consumption