Giuseppe Concialdi
Giuseppe Concialdi

About Me

I'm Giuseppe Concialdi, a Machine Learning Engineer who builds AI systems that work. I take cutting-edge research and turn it into production-ready solutions that solve real problems.

What I Do

I spend my days building AI systems that solve concrete problems. Currently, I’m working on construction technology at Rebar. I work on:

Language and Vision Models

  • Building agents that can understand complex documents and extract meaningful information
  • Developing custom OCR systems that can read technical drawings and floor plans
  • Creating multimodal systems that combine text and visual understanding

Production AI Systems

  • Designing pipelines that process massive documents efficiently
  • Building active learning systems that get smarter over time
  • Implementing CI/CD workflows that keep models accurate and up-to-date

Research Interests

  • AI agent development and autonomous systems
  • Multimodal AI and creative applications
  • Music generation and diffusion models

Background

I started my journey in AI through academic research, where I worked on music generation using diffusion models. That experience taught me the importance of bridging the gap between research breakthroughs and practical applications.

My path led me through various domains, from academic research to ad tech, and now to construction technology. Each step has reinforced my belief that the most impactful AI work happens when you take cutting-edge research and make it work reliably in real-world scenarios.

Technical Approach

I believe in building systems that are not just technically impressive, but actually useful. This means:

  • Focusing on real problems - Not just building cool demos, but solving actual business needs
  • Emphasizing reliability - Making sure systems work consistently in production
  • Continuous improvement - Building feedback loops that make systems better over time
  • Practical deployment - Getting models from research to production efficiently

Writing and Sharing

I write about what I learn while building these systems. My posts cover everything from the technical details of model architectures to the practical challenges of deploying AI at scale. I believe in sharing knowledge—both the successes and the failures—because that’s how we all get better at this craft.

Connect

I’m always interested in connecting with fellow engineers, researchers, and anyone passionate about building AI systems that make a difference. Feel free to reach out: