Saturnalia is a wine advisory platform with a strong USP; it collates and analyses data collected by satellites.
Yet, according to Château Lafite's CEO Jean-Guillaume Prats in a passing comment in a recent interview, the online service is responsible for "misleading consumers". James Lawrence asked Gianni Cristian Iannelli – CEO and president of Ticinum Aerospace, owner of fine wine platform Saturnalia – to outline their mission statement and to answer the charge.
Prior to the interview with Jean-Guillaume Prats, I knew very little about Saturnalia. Can you tell us more about the platform?
I represent a group of five academics who founded a spin-off company called Ticinum Aerospace in 2014. Saturnalia is a branch of our company. It started as an idea developed about three-and-a-half years ago by Daniele De Vecchi, who had great plans in mind for building on his satellite expertise. Out of the five founding members, four of us hold a PhD in electronic engineering; our key focus concerns the analysis of data acquired by satellites.
Since we created Saturnalia, we've obtained financial support from the European Space Agency and the European Commission. We now work with external consultants including agronomists, fine-wine experts and a producer consortium. Our goal is to transform the fine wine world by creating a global "living atlas" with high-quality analysis and insights for any vineyard and wine brand.
"Transform the fine wine world" is quite a grandiose claim – can you explain more about Saturnalia's day to day operation?
The platform continuously ingests a variety of data such as satellite images and weather-related data, from various sources. Then, a network of proprietary artificial intelligence-based models processes the data in order to produce insights about any single fine wine mapped in the system, starting from the vineyards, to wine quality and the current market price.
"Know your wine" is our mantra. Our customers will be able to access a wide variety of information from a single source; data ranging from production to price predictions and fluctuations on the market. Saturnalia can tell you, for example, how much rain in August in a particular year has impacted on the wine you are drinking now, or how much sun exposure a vineyard received in a growing season. We also predict release prices based on our assessment of the vintage quality. The platform is enhancing knowledge and understanding of the fine-wine sector by producing science-based insights. It is a completely new way of exploring the fine wine world.
This is an interesting sales pitch, but can you provide concrete examples of how the platform helps investors and the trade?
Of course. To cite one example: wine investors are able to obtain pre-emptive information on predicted scores and prices, to help them steer their investment decisions. Critics' scores obviously impact on wine prices – this advance knowledge provides the investor a better chance of securing the best deals.
In addition, retailers enjoy a vast pool of early information about the next vintage – very useful to compose an effective procurement list to keep their shelves full with best-selling wines. Collectors benefit from early bargain-picking to a more complex experience where the selection of the most suitable wine for the next fine dining event comes as the results of a journey through the vintage, supported by the platform. Lastly, producers can also use Saturnalia to communicate data to their (potential) customers and keep a close eye on their vineyards and their status, in a dedicated section.
This all sounds useful and impressive – why do you think Jean-Guillaume Prats claimed you "mislead" consumers, and how do you respond to that criticism?
We are aware of JGP's opinion regarding our work, but to read that in such a trenchant and public way has left us surprised, not to mention the lack of any argumentation or clarification whatsoever in the article. This is why I contacted you. We are a science-based proposition, and we have been developing our services for nearly three years now.
However, our experiences in science and research has taught us that criticism helps everyone to improve. For example, we were told by at least one important wine producer that there were inaccuracies in our map of their vineyards. This is a known problem, particularly when the maps aren't provided directly from the growers. It is part of the uncertainty of our model, along with other known and unknown variables. This is something that we have to accept and for which we can't give any absolute guarantee; we know that no model is perfect in itself, and we always welcome whoever helps us to rectify our errors.
But setting aside the issue of maps, JGP's criticism presumably relates to your accuracy regarding predicted scores and vintage quality?
All our score predictions are released several months before critics arrive to taste, and everybody can compare the results in full transparency. Our successes are many: receiving three rounds of competitive funding and the interest of several high-level businesses in the fine-wine industry is surely relevant here. But besides that, we believe that our platform can offer significant value to many actors of the fine--wine business, and that our prediction model does work. On this matter, we have recently published a study on our website, analyzing the prediction accuracy based on verifiable facts. It is available here.
So you totally refute any criticism of the accuracy of your predictive modeling?
We are particularly proud of our monthly price prediction model and analysis for Bordeaux. We are now expanding to many other areas, including Barolo with the fabulous 2016 vintage being released. However, I will concede that predictive modeling can't be perfect – we also have to live with our mistakes. It's part of the game; when you publish your predictions in advance there is no place to hide. Yet our key mistakes have related to the maps – this is impossible to eliminate completely. But we always strive to improve and we welcome constructive and useful criticism.
What have you learnt from your mistakes?
We learnt that mistakes can be turned to strengths if you handle them correctly. In the case of misplaced boundaries, for example, we emerged with a more accurate dataset, ready to produce better insights and to improve the user experience.
Finally, what is your overall impression of the 2019 Bordeaux en primeur campaign?
2019 looks promising for two main reasons: excellent quality and many wines that are released at a discount compared to 2018. Of course, some mentioned that 2018 prices were already quite high, so the discount isn't so great on average. It is, in fact, 2018 wines that might face some difficulties in the market because of comparable quality but higher prices, but we’ll see. Bordeaux's great wines usually show resilience in the medium to long term.
August 08, 2020 at 07:01AM
https://ift.tt/2XFprM1
Bringing the Space Race to the Vineyard - Wine-Searcher
https://ift.tt/31lUVcw
Wine
No comments:
Post a Comment