AI and data hold no secrets for him. On the sidelines of the recent Snowflake World Tour Paris, Frédéric Adet, Data Architecture Practice Leader at Devoteam Data Driven, shares his vision of the dynamics underway in businesses and the prospects offered by these new-generation tools.
Download Now: 50+ AI Reference Cases
[Updated January 2025]
Understanding the Challenges of AI & Data
“Frédéric, could you tell us about the challenges our customers face regarding AI and modern data platforms?”
Among customers in general, there’s a new development: the company’s decision-makers want to impact their business through AI. The impact is twofold: internally, to mobilise their teams’ energy and make things easier to understand, enabling them to work more effectively with data. The next logical step is to generate more value from that data. These decision-makers often lack a clear idea of AI’s capabilities and limitations and how to integrate them into their overall strategy. They are currently justifying a lot of things with this notion.
The idea, which is the whole point of our action, is to share the sense of urgency surrounding AI. It’s important to realise that AI has the potential to deliver unprecedented ROI, but it cannot exist without solid data, technology, business and governance foundations. There can be no AI without data, as the initiatives already undertaken with our customers have shown us. In this approach, we also know that the human element and team acculturation are vital, over and above the technological aspects.
The Importance of Data in AI Initiatives
“Can you tell us more about what Devoteam and our partner Snowflake have to offer?”
At Devoteam, we offer support for data strategy and usage via modern platforms. We also build teams’ technical skills by implementing use case after use case. This approach covers all the human and technical layers necessary to optimise AI, enabling teams to master it. This hands-on method is more effective than a simple need/response approach.
In both human and technical terms, everything is in place for a company to take a fundamentally different approach to its business.
A vast number of things are already possible. With a platform like Snowflake, nothing is impossible. We are currently implementing solutions at our customers’ sites. Our teams utilise their structured data, semi-structured data (from various APIs), and even unstructured data. We then add exploitable semantics to this data for AI utilisation. It is perfectly feasible from a technical point of view.
Unlocking the Power of the Snowflake Platform
“Do you have an example of a concrete use case?”
A primary decision-maker today might want to know how his margins have evolved over the last three months for a specific activity or production line or how his reputation is evolving in a given field. In the best possible conditions, thanks to AI, he simply has to ask questions in natural language to get the answer instantly. We rely on the global informational context, particularly from unstructured data and documents of all types, which are already in the company’s tables and assets. This allows us to have a man-machine dialogue on processes, predictions, and recommendations for new use cases designed with AI.
However, to realise these benefits, decision-makers must translate their impetus and desire into more than just a technical plan. They must acculturate teams to the possibilities. They must also compensate for the debt they owe to using data to get the best out of the old data ‘disruptions’, such as machine learning, prediction, recommendation, deep learning (image and video analysis), etc. These developments have been around for a long time, but these technological leaps didn’t necessarily generate value at the time.
Download Now: 50+ AI Reference Cases
[Updated January 2025]
The Future of AI & Data
“How do you see things developing in the near future?”
Life will be a lot simpler for all company employees in the future. They’ll be able to be much more proactive and autonomous in helping their business unit or area of activity shine. That’s what modern data platforms are all about, like Snowflake, our technology partner. As a former Data Scientist, I know that data specialists, such as Data Scientists, Data Engineers, and even Data Analysts, can now work across the whole platform, from data ingestion to data transformation, via governance. Everything is made very simple and easily deployable, with a time-to-market that defies all competition.
With a platform like Snowflake, AI makes it easier to understand, measure and cross-analyse company data. Everyone in the company is empowered to experiment with their value proposition ideas.
Thanks to this tool, an idea from a good decision-maker or team leader in a specific sector can quickly become concrete and validated by the various departments. With these platforms, you can experiment with new ideas. They can be validated, industrialised, published and made available to company staff and external stakeholders, such as customers and web users. I’d say that today, a good idea can be implemented in two hours to three days by a handful of people, including a validator. In contrast, in the past, many different roles and stages were involved: funding, sponsorship, project manager, service provider, etc. And it used to take six months!
The “No AI Without Data” Principle
“What is the meaning of “no AI without data”?”
This means good AI depends on properly prepared, clean data with specific semantics and context. Teams must qualify and catalogue the data to quickly understand its use, quality, and potential value. You cannot build relevant and reliable AI on low-quality data.
Devoteam helps you lead the (gen)AI revolution
Partner with Devoteam to access experienced AI consultants and the best AI technologies for tailored solutions that maximise your return on investment. With over 1,000 certified AI Consultants and over 300 successful AI projects, we have the expertise to meet your unique needs.