• KS meets Jacopo Berti, CEO of Rozes. Revolutionising Risk Management with AI

KS meets Jacopo Berti, CEO of Rozes. Revolutionising Risk Management with AI

April 29, 2024


Knowledge Share

In the vast landscape of emerging and promising start-ups, Rozes is making a national and international impact in the field of risk management.

A spin-off of the University of Padua, Rozes is the most visualised start-up of the month of March on the KS 2.0 platform. The driving force behind the development of the project is Jacopo Berti, Managing Director of Rozes, who, together with the project team, has turned an innovative idea into a tangible reality.

Rozes distinguishes itself by offering security services to entrepreneurs, companies and public administrations through the production of reports and the monitoring of budgets, all thanks to Artificial Intelligence. In an increasingly complex Italian and international context, the subject of Corporate Intelligence and Risk Intelligence is gaining unprecedented centrality.

Jacopo Berti recounts Rozes' journey and underlines the crucial importance of this company in the fight against organised financial crime and the protection of entrepreneurs, companies and public funds. The Artificial Intelligence algorithm developed by Prof. Antonio Parbonetti - Pro-Rector of Università di Padova - and his research team is able to identify similarities with criminal enterprises by analysing financial statements, thus enabling the prevention of reputational, legal and industrial risks.

Rozes - team building

Scenario and Panorama - Rozes offers security services to entrepreneurs, companies and public administrations by producing reports and monitoring budgets using AI. What is the current risk management scenario in Italy?

Today, the subject of Corporate Intelligence and Risk Intelligence is certainly becoming more and more central, for the simple reason that the national and international scenario is becoming increasingly complex.

We resolve this complexity in the main issues we deal with - i.e. the fight against organised financial crime, all the companies linked to it and all the geopolitical risk scenarios for entrepreneurs, companies and public administrations - by offering solutions that manage to protect entrepreneurs, companies and public funds, but also to help law enforcement agencies (as far as we can). Not only that, but our contribution is to make citizens and businesses safer.

When and how was Rozes born?

Rozes was born from the study and scientific activity of Prof. Antonio Parbonetti - Pro-Rector of the University of Padua - who, together with his research team, invented this artificial intelligence algorithm that is able to detect the similarity of a company's balance sheet to so-called criminal companies.

Criminal companies are defined as companies that have been convicted of 416 bis: mafia-type criminal association and all related crimes, such as false invoicing, money laundering, fraudulent bankruptcy, etc.

This ability of ours to go and identify, thanks to the analysis of financial statements, the similarities of companies with patterns of criminal reality, allows us to go and identify in real time the potential risks that lurk within customers for banks, suppliers for large companies, to name but a few. In this way, the entire supply chain is protected against reputational, legal and, of course, industrial risks, but above all against the activities of criminals.


Artificial intelligence is a hot topic these days. It is being talked about in very different situations and seems to be the new frontier in approaching innovation and beyond. But you have been using AI for some time. Are you somewhat ahead of the curve in your use of artificial intelligence?

If you think about it, we have been working on it since 2018. The first drafts of the algorithm were born in that year, when the work of the research team also started, so we have been studying it for many years, considering that we come from economics with a mix of mathematics and data science.

AI is the 'buzzword' of the moment, but you should never confuse what kind of 'intelligence' you mean. We specialise in machine learning, which is all about financial classifiers. Then, of course, we use generative AI to implement activities. But it is very, very important to say that it is a project that is the result of years of work and scientific research with publications in international journals, because that is our methodology. Everything comes from research: scientific research, publication of 'papers' and then the spin-off does the technology transfer, so the third mission. Rozes is based on the inventiveness and scientific research carried out at the university.

Do you think that the recognition of its importance, both as a 'problem' and as a 'solution', at a national level, is an awareness that has existed for some time, or is there a kind of awareness that has been dictated by recent times (and trends)?

I think it is a matter of journalistic sensationalism. There is a very normal learning curve and recognition of new methodologies, but this is accompanied by a sensationalism that wants to bring to the fore something that is described as either a catastrophe or a panacea for humanity.

In a world of polarisation, where people are always trying to be either 'for' or 'against' something, it is important to remember that there are statistical methods that regulate everything. They are, in fact, statistical methods that help predict what they are supposed to predict. No more, no less.

Then I am also convinced that the name 'Artificial Intelligence' carries with it considerable media disruption. Artificial intelligence': let us say that it arouses in us the idea of the machine that wants to overpower man, which is a completely false idea.


How do you approach the potential stakeholders of the services you propose? You include financial institutions, companies of a certain size. And what kind of feedback do you get?

There is a very different approach between different end users. There are those who use it for anti-money laundering control activities and third party assessments. And that is very important because it helps us to understand how different stakeholders approach the service.

The key thing that we have noticed in recent years is the growing interest in the security of the citizen or the entrepreneur. There is a very, very high awareness and this is a cultural key because then the technological tools are nothing more than an application of what we want and what we think. And in the end, it is always thought and culture that guides action.

What is your business model? Can the Rozes Index have an economic and strategic impact on the start-up?

The Rozes Index is an index that, from 0 to 1, highlights the similarity of a company to criminal type realities. This is where it all starts, and from this index a series of risk intelligence services, platforms and technologies are born, applied to anti-money laundering, third party management, tendering, procurement and other areas.

Remember that these services are provided to companies through platforms that we produce and that are managed autonomously by our clients, together with a consulting activity of support and training for all that is the knowledge of new methodologies and new applications to always use innovative technologies at the state of the art for the fight against crime and especially for the security of their company.

As a business model, we are a SaaS - Software As A Service - and we make our platform indexes available to our customers. Then we provide training, consultancy and change management, i.e. change in use and accompanying services.

In the 'Looking for' section of your KS 2.0 factsheet, 'investment funds' appears with an 'early stage' funding stage. How do you reconcile these two elements for the support and development of Rozes?

Today we are not really looking for capital and if we were to open this option tomorrow it would be 'patient capital'. It is capital that accompanies and grows with an industrial, financial and strategic partner, which is different from a venture capital approach.

For us, there is a very ethical vision of the impact of our work. We do not want to be 'sector multipliers' and then sell the company on. We are here to create real and substantial change, and to generate impact in a way that is currently at odds with risky methods.


Are you also approaching an internationalisation process?

We have some clients from abroad. They are multinationals that are also present in Italy and they want to make sure that they operate transparently and safely.

In terms of approach, I have to say that we Italians are very prepared and careful. In terms of data management, where we have a very high capacity and where we have done a great job at the level of legislation (which is very solid), there is a good attitude. However, this should be backed up by an equal cultural solidity that is not only oriented towards business but also towards the betterment of society. Many companies have this, and fortunately it is growing day by day.

Risk management and NRRP - National Recovery and Resilience Plan . Have you noticed any changes as a result of the plan? Do you feel that more opportunities or supportive scenarios have opened up?

Certainly the NRRP is a great accelerator for the whole country and the data says so. Unfortunately, there is a growing theme of the risk of infiltration: so much money, increased 'appetite'.

I have to say that we are very happy to play our part, because companies are really sensitive to this. They realise that risk management is a key issue to ensure business growth.  When and how did you realise that your project was working?

We realised when the first algorithms were circulated in the university about thousands and thousands of companies that we knew had already been convicted. Our algorithm didn't know that, and it caught us in a very timely manner, so much so that it earned us publications in international accounting journals, proving that we had invented something that is still unique in the world and raises the bar worldwide in terms of research and application of AI for these purposes.

For more information on 'Rozes: Rozes (knowledge-share.eu)

Share Article