Artificial Intelligence (AI) has quietly embedded itself into our daily lives. It helps us build shopping lists, find answers to both simple and complex questions, and even draft messages in seconds. For many of us, using AI has become second nature.
And large enterprises have clearly caught on: this technology is a powerful lever for both productivity and innovation. But between media hype and operational reality, a key question remains: How can we turn the excitement around AI into robust, relevant, and long-lasting use cases?
That’s exactly the challenge major corporations are now facing. To shed light on it, we spoke with Benjamin L’hyver, Chief Data Science Officer, and Guillaume Jaeger, Director of Consulting & Transformation at Eulidia, a consulting and technical expertise firm specialized in Data and AI since 2008. They shared their on-the-ground insights into the real needs of large organizations, along with best practices to adopt.
A thriving market, in search of focus and operational clarity
AI is everywhere today, with teams across departments running countless experiments, an encouraging sign of progress. But true value only emerges when these initiatives are transformed into robust, long-lasting solutions.
Behind the growing excitement around AI lie new, complex challenges : How can companies ensure project relevance, guarantee compliance, and avoid scattered, disconnected initiatives across departments? For large organizations, the key is now to channel this energy, align it with business priorities, and embed it into a broader, structured change management approach.
Today, leadership teams are looking less for creativity at all costs, and more for robustness: solutions that are reliable, fast to deploy, compliant by design, and seamlessly integrated into existing processes.
As a result, three major priorities emerge :
Upskill and educate teams to make AI projects more sustainable and ensure long-term adoption.
Standardize practices to improve compliance and reduce risk.
Rationalize AI use cases to cut costs and unify methods across the organization.
Examples of AI Use Cases in Large Enterprises
While the rise of generative AI has broadened the horizon of possibilities, the use cases currently being deployed within large enterprises remain highly pragmatic, with a strong focus on operational value. Among the most common examples:
Search engines that understand natural language queries.
Domain-specific assistants for technical and back-office functions.
Automation of responses to Requests for Proposals (RFPs).
Narrative reporting.
On the technology side, several standards are emerging, including platforms like Snowflake, Dataiku, and Databricks, as well as mainstream LLMs from providers such as OpenAI, Mistral AI, and Google. No-code tools like N8N are also gaining traction for orchestrating workflows with minimal technical input.
A new playing field, with fundamentals that remain unchanged
According to Guillaume Jaeger, every AI initiative, like any data project, relies on three complementary layers.
The first layer is the business layer. On one side, end users express their operational needs; on the other, business leaders define priorities and shape the roadmap.
The second layer is the technical layer, made up of data scientists, data engineers, data architects, and analysts. These teams design the solutions and build the technological foundations.
Between the two lies a crucial layer: the Product Owners and Product Managers. They act as conductors, ensuring alignment between business and tech. They translate business needs into concrete use cases and ensure each project delivers real, measurable business value.
Without a clear alignment between business teams, Product Owners/Managers, and technical experts, AI projects risk becoming fragmented, experimental, or disconnected from real business value. But when all components work together, AI evolves into a robust, relevant solution, fully adopted by users.
And contrary to popular belief, AI hasn’t radically changed the core skills required. “What makes a great Product Owner hasn’t changed with AI,” explains Guillaume Jaeger. What still makes the difference is:
Clear communication, both written and spoken
The ability to explain complex technical concepts in simple, accessible terms
A strong grasp of business priorities and strategic goals
The foresight to anticipate how the application will be used over time
Efficient and pragmatic project management
That said, AI brings one new dimension to the forefront: the importance of UX/UI. AI tools are often designed for users who aren’t necessarily familiar with technical solutions, which is precisely what makes them so powerful. However, this also means that the interface becomes a critical factor for adoption.
Choosing the right use case for a successful AI project
More than the technology itself, the success of an AI project relies on a solid, structured approach.
At Eulidia, that begins with identifying and prioritizing the right use cases. This is done using an evaluation framework made up of 54 criteria, grouped into 5 pillars, and ultimately synthesized into two key dimensions: value and feasibility, to focus efforts on the most impactful initiatives.
Another key pillar is user adoption, driven by a sponsor from the end-user side, the product owner—who is involved from day one.
This person plays a central role in gathering feedback and ensuring the solution evolves continuously to meet user needs.
When it comes to ROI, Eulidia promotes a “humble and pragmatic” approach, explains Guillaume Jaeger.
Quantifiable results—like user adoption or time saved—are tracked. But it remains difficult to measure everything, especially when solutions support cross-functional teams.
Still, several tangible successes have already been observed:
Several weeks saved on RFP responses
Greater autonomy for business units and analysts
Easier access to complex knowledge bases for end users
Faster and more efficient execution of complex industrial interventions
Reduced pressure on support teams through increased self-service adoption
Accelerated onboarding of new employees via AI-powered assistants
Minimized impact when key personnel leave the organization
There are some points of caution to keep in mind. While maintenance costs are typically anticipated during the design phase, usage-based costs especially those tied to query volumes, remain much harder to control.
In addition, cybersecurity and data privacy are critical concerns. These must be addressed upstream, from the very beginning of the project, to ensure the solution is secure by design and compliant with enterprise-grade standards.
Toward a New Balance Between IT and Operational Teams
AI is not only transforming processes—it’s also reshaping project governance. As Benjamin L’hyver points out : “With GenAI, it’s no longer just the IT department driving projects, business teams are now actively seeking out technical expertise.”
For large enterprises, this marks a true shift. The days when data initiatives were siloed within IT are over. Today, business units are directly involved in defining and driving projects forward. This evolution requires new ways of working, where every stakeholder, IT, business, and cross-functional teams, plays an active and clearly defined role.
To support this transition, LittleBig Connection enables large organizations to quickly access the expertise they need, by creating a direct, transparent link with freelancers and specialized consulting firms like Eulidia. A powerful lever to turn AI into real-world projects that deliver measurable business value.
