Frequently Asked Questions

TORA NO AI is an open source solutions lab, incorporated as an Italian Benefit Corporation (Società Benefit) with €90,000 share capital fully paid. We take tech problems the market won't solve, solve them, and publish everything — source code, technical whitepapers, documentation. We work on distributed technologies because critical infrastructure shouldn't depend on a single entity. We release everything open source because whoever sets the standard sets the market — and we want that standard to be open.

A Società Benefit (Benefit Corporation) is a legal form introduced in Italy by Law 208/2015 (Stability Law 2016, Art. 1, sections 376-384). It requires the company to pursue, beyond profit, common benefit purposes toward people, communities, territories, and the environment. TORA NO AI publishes an annual Impact Report to account for the impact generated.

Distributed AI is an approach where artificial intelligence models run on heterogeneous infrastructure — from industrial servers to edge devices — without depending on a single central point. TORA NO AI designs algorithms that work this way to ensure resilience, data sovereignty, and elimination of single points of failure.

Federated learning is a technique where AI models are trained directly on the devices where data resides, without ever centralizing it. Only updated model parameters are shared, not raw data. TORA NO AI uses this approach to protect privacy and reduce the transfer of sensitive data.

Vector databases are storage and search systems based on semantic meaning, not exact keyword matching. Each piece of data is converted into a numerical vector representing its conceptual content. TORA NO AI uses them to build semantic memories within its distributed AI systems.

TORA NO AI welcomes collaboration proposals on research projects, technology partnerships, and questions about published whitepapers. To propose a collaboration, use the contact form and indicate the technical context. Every message is read directly by those who work on the projects.

Yes. TORA NO AI publishes 100% of its source code as open source. Every completed project includes a technical whitepaper, synthetic datasets, and a public repository. Visit the Projects section to explore ongoing research.

The Impact Report is the annual document that TORA NO AI publishes as a Benefit Corporation, as required by Italian Law 208/2015. The assessment follows B Lab standards. It accounts for actions taken and results achieved with respect to the reference UN SDGs (4, 7, 8, 9, 11, 12, 16, 17). Read the full Impact Report.

TORA NO AI integrates environmental impact assessment into every research project. Every line of code has an energy cost: we optimize for computational efficiency, leverage existing hardware, and measure associated emissions. Results are documented in the annual Impact Report.