75+ AI agents.
One document.
Zero astronauts.

PAK4L is named after K'inich Janaab' Pakal, the Maya king of Palenque who ruled for nearly seventy years in the 7th century. His sarcophagus lid was so enigmatic that for decades UFO enthusiasts were convinced it depicted an astronaut piloting a spacecraft. It actually represents his journey into the underworld.

That's exactly what happens with documents: everyone sees astronauts. Almost nobody bothers to study the cosmology.

The "4L" stands for Key 4 Logic — the core principle behind deep, evidence-first review.

Alberto Giovanni Gerli — entrepreneur, CEO, and innovator. A career built across business development, advanced technologies, complex processes, and artificial intelligence applied to real work.

He has built experience in contexts where precision, document quality, compliance, and decision-making truly matter: industry, professional lighting, smart cities, public procurement, technical documentation, and regulated environments.

Over the years he has led companies, developed products, built businesses, and worked on processes where value doesn't come from theory — but from the ability to transform complexity, risk, and information into concrete actions.

Alongside his entrepreneurial and managerial path, he spent three years as a researcher at the Università degli Studi di Milano, Department of Clinical Sciences and Community Health.

In that context he published several scientific works on artificial intelligence, the limits of generative models, chatbot evaluation, the use of ChatGPT and LLMs in scientific writing, and the development of predictive models on epidemiological data — including studies related to Covid-19. This experience gave him a rigorous, critical, evidence-oriented approach that he now brings into product development.

PAK4L was born from the intersection of these two paths: on one side, AI research; on the other, years of work on complex documents, decision processes, reviews, risk, and accountability. He wanted to build a platform that doesn't just "read documents" — but analyzes them in depth, compares different perspectives, applies specialized logic, makes the process more traceable, and helps people and organizations make better decisions.

His goal with PAK4L is clear: bring artificial intelligence beyond the demo stage and turn it into a serious, structured, useful tool — one that increases quality and control in the most critical document workflows.

Built from 9 PM to 4 AM, every night, for months. Not because it was required. Because it couldn't wait.