Sponsored byMistral AI
A multitude of organizations jumped into generative AI, only to find that pilots fell short of providing value. Presently, firms seek measurable results—but what strategies will ensure success?
At Mistral AI, we collaborate with leading global industries to co-create customized AI solutions that address their toughest challenges. Whether enhancing CX productivity with Cisco, developing a smarter vehicle with Stellantis, or speeding up product innovation with ASML, we commence with open frontier models and tailor AI systems to make an impact for each organization’s distinct challenges and objectives.
Our approach begins with pinpointing an iconic use case, the cornerstone for AI transformation that lays the groundwork for future AI initiatives. Selecting the appropriate use case can signify the difference between meaningful transformation and continuous tinkering and trials.
Pinpointing an iconic use case
Mistral AI identifies four essential criteria when evaluating a use case: strategic, urgent, impactful, and feasible.
Initially, the use case needs to possess strategic significance, tackling a key business process or enabling a groundbreaking new feature. It should transcend mere optimization; it must be transformative. The use case should be sufficiently strategic to engage an organization’s C-suite and board of directors.
For instance, internal HR chatbots are convenient but are simple solutions that do not foster new innovations or possibilities. Conversely, consider an external banking assistant that not only addresses inquiries but also facilitates actions like blocking a card, executing trades, and recommending upsell/cross-sell prospects. This exemplifies how a customer support chatbot is elevated to a strategic, revenue-generating asset.
Secondly, the most suitable use case to pursue must be acutely urgent and remedy a crucial business issue that stakeholders care about immediately. This endeavor will require employees’ time—it must be significant enough to warrant that time expenditure. Moreover, it needs to assist business users in overcoming pressing challenges.
Thirdly, the use case must be pragmatic and impactful. From day one, our mutual objective with customers is to implement a solution in a real-world production context to enable testing with actual users and gather insights. Numerous AI prototypes end up discarded as flashy demonstrations that are inadequate for customer interaction, often lacking frameworks for evaluation and enhancement. We partner with clients to ensure prototypes are sufficiently stable for release and that they have the essential support and governance structures in place.
Finally, the ideal use case should be feasible. There may be various urgent initiatives, but selecting one that can yield quick returns facilitates the momentum required to continue and expand.
This entails searching for a project that can be operational within three months—and a prototype can be active within a few weeks. Rapidly deploying a prototype for end users is crucial to collect feedback and ensure the project remains on course, making adjustments as necessary.
Where use cases underperform
Organizations are intricate, and the way forward is typically not clear. To sift through all possibilities and identify the right initial use case, Mistral AI conducts workshops with clients, closely collaborating with subject-matter experts and end users.
Representatives from assorted functions will demonstrate their processes and discuss potential business cases for a first use case—and collaboratively, we determine the standout option. Here are examples of project types that do not qualify.
Moonshots: Ambitious undertakings that inspire leadership but fail to chart a quick ROI path. While these projects can be strategic and urgent, they seldom meet feasibility and impact criteria.
Future investments: Long-range endeavors that can delay. Although these initiatives can be strategic and feasible, they rarely satisfy urgency and impact needs.
Tactical fixes: Immediate-response projects that resolve short-term issues but do not create substantial progress. While such situations can be urgent and feasible, they often fall short on strategy and impact.
Quick wins: Advantages in generating momentum, but not transformative. While they can be impactful and feasible, they seldom meet strategy and urgency standards.
Blue sky ideas: Projects that could transform but need maturity to be practical. Although they can have strategic and impactful potential, they often lack urgency and feasibility criteria.
Hero projects: High-pressure initiatives that lack executive support or realistic timelines. While they can be urgent and impactful, they typically lack strategy and feasibility.
Transitioning from use case to deployment
After identifying a well-defined and strategically sound use case ready for development, the next phase is validation. This entails conducting initial data exploration and mapping, determining pilot infrastructure, and selecting a target deployment scenario.
This phase also includes finalizing a draft pilot scope, determining participants for the proof of concept, and establishing a governance framework.
Upon completion, we move into the construction phase. Organizations collaborating with Mistral engage with our in-house applied AI scientists who develop our frontier models. We jointly design, build, and launch the initial solution.
Throughout this stage, we emphasize co-creation, enabling us to pass on knowledge and skills to the organizations we assist. As a result, they can maintain independence far into the future. The outcome of this stage is an operational AI solution with empowered teams capable of autonomous action and innovation.
The initial step is crucial
After achieving the first success, it’s vital to leverage the momentum and insights gained from the iconic use case to discover more high-value AI solutions for implementation. Success is defined as having a scalable AI transformation model with multiple high-value solutions integrated throughout the organization.
However, none of this could occur without successfully pinpointing that initial iconic use case. This initial step transcends merely selecting a project—it is about laying the groundwork for the entire AI transformation.
It distinguishes between disjointed experiments and a strategically aligned, scalable journey toward impact. At Mistral AI, we have witnessed how this methodology unlocks measurable value, aligns stakeholders, and fosters momentum for future advancements.
The route to AI success begins with a single, thoughtfully selected use case: one that is bold enough to motivate, urgent enough to necessitate action, and pragmatic enough to achieve results.
This content was produced by Mistral AI. It was not written by MIT Technology Review’s editorial staff.
