On October 29 and 30, Deix participated in SMAU Milan, a key event for innovation in Italy, hosted at FieraMilanoCity. With the support of the Lombardy Region, we were present with our booth alongside other local excellence, representing the best of Lombard innovation.

The experience at SMAU allowed us to meet with prominent corporations interested in our Mathematical Intelligence solutions and our expertise in AI and mathematical modeling. These meetings opened up collaboration opportunities with organizations that share our goal: accelerating companies' digital transformation in an ethical, sustainable, and transparent way.

Participating in the workshops dedicated to innovation and artificial intelligence was fundamental for us to engage with emerging trends and, above all, with the main pain points of companies. Among the critical challenges discussed, the adoption of AI tools within companies stood out: bringing innovation also means overcoming cultural and structural barriers to ensure the effective and advantageous implementation of advanced technologies.

We also had the opportunity to participate in two speed-pitching sessions, where we presented our solutions to an audience of industry experts and decision-makers.


The two days at SMAU were intense but extremely productive, thanks to the entire Deix team—Jacopo, Fabio, Rossella, and Francesca—and our partners at Ottozerouno, Luca and Marco, for the great work done!

Deix at SMAU Milan 2024

We are proud to announce that Deix has received the DigiSafeWell Award.

This recognition is an important affirmation of Deix's commitment to creating advanced technological solutions for workplace safety, well-being, and sustainability. Our platform, [Varda-AI](https://deix.ai/en/solutions/products/varda-ai), awarded for its ability to monitor production environments and reduce risks through Artificial Intelligence, is an example of how we aim to transform corporate safety into an operational advantage.

Thank you to FAST, Enterprise Europe Network - Confindustria Lombardia, and all the organizers for this opportunity. Congratulations as well to the other winners who are driving incredible innovations in their respective fields!

Deix Wins the DigiSafeWell Award

From October 21 to 23, we were in Catania at the prestigious Monastery of the Benedictines to participate in the Research Exchange Forum organized by the Emblema Foundation. An unmissable event that brings together companies, startups, investors, universities, and research centers with the goal of fostering innovation and collaboration among these players.

During our plenary pitch, we had the opportunity to present our unique vision of innovation, based on **Mathematical Intelligence**, and shared some of our latest research projects in the field of Large Language Models (LLM) and Generative AI, which aim to make these tools more accessible and interpretable. These projects reflect our commitment to the democratization of AI and the creation of models that are not mere "black boxes" but transparent and ethical systems.

Returning from this experience, we bring back a network of partners with whom we can work to continue pushing the boundaries of innovation and applied technology, as well as a wealth of insights and valuable feedback to refine our proposals in the industrial sector and digital transformation.

Deix at the Research Exchange 2024

On July 15th, we had the pleasure of hosting a plenary day with our scientific committee at Deix's headquarters in Milan. It was a special occasion, not only to meet in person but also to discuss key topics for the company's future.

Professors from Rome and Florence actively participated in the discussion, offering valuable insights on many fronts. We covered updates on our current projects and those involving new clients, highlighting the progress made and future milestones. Among the main topics was Deix’s new direction towards Mathematical Intelligence, which combines mathematical models and artificial intelligence to provide unique and differentiating solutions. This new positioning represents an important step in communicating our external identity and the values that make us unique.

In addition to this, we shared updates on business development, participation in public tenders, and the excellent results achieved in communication activities, particularly regarding the growth of our presence on LinkedIn over the past six months. We also had a constructive discussion about marketing activities, including the creation of promotional materials and participation in industry events such as AI Week in Rimini and We Make Future in Bologna.

Finally, we reflected together on the outcomes of the Deix Reflections event in May, a brainstorming day that allowed us to focus on how to further improve our internal processes.

Meeting in person was essential: moments like these allow us to align on future goals, strengthen team cohesion, and continue building Deix’s future with enthusiasm.

The future of Deix in our plenary meeting

Mathematical Intelligence and New Goals

Between June 24 and 28, the DFOS 24 (Derivative-Free Optimization Symposium 2024) conference was held in Padua, an event of international significance in the field of derivative-free optimization. The conference featured the participation of global experts who shared their latest research and innovations. Deix was a sponsor of the event, and Giampaolo Liuzzi, a founding partner of Deix and a member of our scientific committee, played a prominent role as a member of the organizing committee.

**The Importance of DFOS 24**

From June 26 to 28, Padua became the focal point for the academic and industrial communities interested in derivative-free optimization. Derivative-Free Optimization (DFO) comprises techniques for optimizing complex objective functions without the need to compute derivatives. This approach is extremely useful in many industrial applications where the objective function is difficult to model or where derivatives are hard or impossible to calculate. Examples include the optimization of manufacturing processes, the design of complex engineering systems, and the resolution of problems with noisy or incomplete data.

**Contributions from Deix and Giampaolo Liuzzi**

Giampaolo Liuzzi presented several cutting-edge research papers on derivative-free optimization methods, demonstrating how these can be applied in practical contexts. For instance, among the projects he is responsible for, the BIODRONES project is one of the most interesting. This project aims to develop autonomous underwater vehicles with biology-inspired shapes and functions, using advanced simulation and optimization algorithms for efficient design.

At Deix, the fusion of academic research and practical applications is the cornerstone of our success. Our mission is to provide innovative, tailor-made solutions to our clients, enhancing their market competitiveness. We are proud to sponsor and actively participate in events like DFOS 24, which allow us to stay abreast of the latest trends and innovations in our field.

Participating in DFOS 24 has allowed Deix to consolidate its position as a leader in mathematical optimization and to strengthen its network of academic and industrial collaborations. We will continue to work with passion and dedication to drive innovation and develop solutions that truly make a difference.

DFOS 24 Conference in Padua

Deix had the pleasure of participating in the WMF 2024 at the "KNOWLEDGE SHARE area - Where Innovation Happens". We had the opportunity to meet so many innovative startups and cutting-edge companies, all united by a great interest in the world of innovation.

At our booth, and during our pitch, we presented our approach of Mathematical Intelligence. Thanks to the integration of mathematical modeling and Artificial Intelligence algorithms, we showcased the unique potential of our solutions, capable of bringing innovation and tangible improvements to business processes.

We had stimulating discussions and laid the groundwork for possible future collaborations. A special thanks to everyone who visited our stand. We are excited about the future prospects and look forward to continuing this journey towards innovation together with you.

Deix was at the WMF event in Bologna

We Make Future 2024

In the digital era, Artificial Intelligence (AI) plays an increasingly central role in our daily lives. However, many of its internal mechanisms remain shrouded in mystery, encapsulated in what experts refer to as "black box" models. These complex systems, while extremely powerful, often lack something fundamental: interpretability.

Recent studies are paving the way to make these models not only more effective but also more understandable. Through the use of mathematical optimization methods, researchers are developing techniques to "open up" these black boxes, making AI more transparent and accessible to everyone.

Mathematical optimization, a branch of applied mathematics that deals with finding the best possible outcome within a given set of parameters, is proving to be crucial in this process. By employing advanced algorithms, it is possible to analyze and modify AI models to make them not only more efficient but also easier to interpret.

A recent study, titled "Supervised feature compression based on counterfactual analysis" (Piccialli, Veronica, Dolores Romero Morales, 2023), marks a significant step in this direction. The main goal of this work is to improve transparency and understandability using a technique called "counterfactual analysis." Counterfactual explanations seek to answer questions such as: "What would need to change in a given input to achieve a different outcome?" For example, if a model determines that a customer is not eligible for a loan, counterfactual analysis could indicate what changes in the customer's data could lead to a positive decision.

The article proposes a method to utilize the information obtained from counterfactual explanations to construct a decision tree, which is a type of machine learning model much easier to understand because it can be visualized as a series of binary questions and answers (yes/no). The created tree seeks to mimic the behavior of the original model but in a much more transparent form. The significant advantage of this approach is that it makes machine learning models more transparent and trustworthy, especially in sensitive sectors such as healthcare and finance, where understanding the "why" behind a decision can be crucial.

Mathematical optimization plays a crucial role in the article, being the key to implementing and improving the proposed methodology of supervised feature compression based on counterfactual analysis, allowing the authors to transition from an opaque and complex machine learning model to one that is more interpretable and transparent.

Numerical results on real-world datasets confirm the effectiveness of this approach, not only in terms of accuracy but also in reducing the complexity of the model. This study opens up new perspectives for making AI more transparent and trustworthy, a crucial step towards the ethical and responsible adoption of these technologies in our daily lives.

Deix stands out in the technological landscape thanks to its unique ability to integrate advanced expertise in applied mathematics, optimization, and numerical simulation with the most modern techniques of artificial intelligence and machine learning. This synergy creates a powerful combination that we call "Mathematical Intelligence." This approach not only improves the understanding and ethics of AI models but also significantly increases their reliability and accuracy.

Deix's experts, with solid academic roots and extensive experience in the fields of mathematical optimization and AI, are developing innovative methodologies to "decrypt" complex AI models. This pioneering work not only opens up new avenues in building transparent AI models but also sets a new standard for the responsible and conscious use of AI in critical contexts such as health, safety, and finance.

Deix's commitment to making artificial intelligence more accessible and understandable goes beyond mere technological improvement. The goal is to build solid trust between users and advanced technologies. By facilitating a greater understanding of how decisions are made by AI systems, Deix not only aims to democratize AI technology but also to strengthen the bond of trust with society. This transparency is essential for true accountability and social acceptance of advanced technologies.

Deix's commitment to **"Mathematical Intelligence"** is not just a matter of efficiency but an ethical choice aimed at ensuring that the technologies we build improve people's lives sustainably and fairly.

In conclusion, as we continue to navigate the complex world of AI, it is essential that we do not lose sight of the importance of transparency and interpretability. Research efforts in this field not only enhance our technological capabilities but also strengthen the bond of trust between humanity and the machines we choose to build.

Making "black box" models understandable

Mathematics at the service of Artificial Intelligence

Deix's recent participation in Rimini's AI Week proved to be an extraordinary experience for our team, represented by Jacopo Corno, Head of Business Solutions, Mattia Satta, Lead Data Scientist, and our OTTOZEROUNO partners. The event provided an opportunity to meet numerous companies, including potential clients and partners, fostering fruitful exchanges and future collaborations.

Throughout the event, we attended various speeches and panels, making sure we are always on top of emerging market trends, discovering new technologies, and sharing industry best practices. A special thanks goes to Giacinto and Pasquale, the organizers, for their excellent work in making AI Week one of the most significant events in the Italian innovation landscape.

The choice of Rimini as the location further enhanced the event, with its welcoming and stimulating atmosphere. This experience not only reinforced our commitment to technological advancement but also expanded our network of strategic contacts in the AI sector.

Articles and news