Choosing your first AI agent: a look at the CNOSF

In a rapidly changing world, marked by the rise of artificial intelligence (AI), organizations are facing unprecedented challenges. How can they optimize their operations while maintaining the human connection? The example of the CNOSF (French National Olympic and Sports Committee) illustrates the possibilities offered by AI agents in nonprofit organizations. During a recent webinar, Salesforce revealed how to choose your first AI agent, the essential steps to follow, and the concrete benefits achieved. The Necessity of Artificial Intelligence for Modern Organizations With 300 million individuals using generative AI every week, it is clear that this technology can no longer be ignored. Many companies, including nonprofit organizations, understand that this innovation is imperative to remain competitive.

According to recent studies, 70% of employees believe that AI will transform more than a third of their work within two years. Furthermore, 41% of organizations’ time is often wasted on repetitive, low-value tasks, highlighting the importance of automating certain functions.

The Benefits of Artificial Intelligence in the Nonprofit Sector Using AI agents can help nonprofits uncover untapped data and better understand their audience. This can result in: In-depth knowledge of donors and benefactors.

Valuable assistance to your members, as the CNOSF does.

Improvements across all organizational processes.

AI as a lever for automation

AI agents automate a variety of processes. This automation can be particularly beneficial in organizations lacking human resources. By facilitating information searches, these agents can help significantly reduce wasted time and direct efforts toward higher-value tasks. Problem AI Solution
Benefits Repetitive TasksAutomation via AI Agents
Time Saving

Difficulty Finding Information

Agent Serving as a One-Stop Shop

Reduced Search Time	Limited Data Access	AI Data Analysis
Better Member Insights	Agentforce: A Tool for Nonprofit Organizations	Salesforce recently introduced
Agentforce	, an innovative product that allows nonprofit organizations to design AI agents integrated into their digital infrastructure. This tool facilitates the creation of agents capable of understanding and processing written requests. This has piqued the interest of several organizations, including the	CNOSF
. To embark on this adventure, several steps must be followed:	Determine the agent’s ultimate objective, specifying the tasks to be performed and the required data.	Evaluate and prioritize use cases by considering impacts and feasibility.

An intuitive interface for everyone

One of Agentforce’s strengths lies in its simplicity. Its administration interface requires no particular technical expertise. Users can define: How the agent should formulate their responses.The discussion topics they can address. The language in which they should interact.This accessibility allows all associations to benefit from AI, regardless of their level of technical expertise.

CNOSF’s practical experience with its AI agent

CNOSF
took the plunge by integrating an AI agent into its infrastructure in recent years. After revamping its member relationship management system, the association set out to create a resource center to facilitate access to information for its members. But how exactly was the agent designed and implemented?

Clément Migeon, Director of Information Systems at CNOSF, explains that their approach refocused on the added value an AI agent could bring. Instead of remaining confined to a simple information bot, they wanted this agent to free up time for their small team so they could focus on higher-value tasks.

The process of creating CNOSF’s first AI agent

The development time for this agent was relatively short, taking only four weeks. This speed was the result of intense collaboration with Salesforce. “We learned not to fear the unknown,” emphasizes Migeon, as this project represented a breakthrough in technology use within the association.
The initial results are positive, but the team also identified several areas for improvement, including structuring the available information for better use by the agent.
Development Phases

Duration

Objectives

Needs Analysis 1 week Define Objectives

Agent Design 2 weeksProgramming via Agentforce

Testing and Adjustments

1 week

Verification of Responses and Access

Future Prospects for AI Agents in the Nonprofit Sector	With the encouraging results obtained with their first AI agent, the CNOSF is now considering expanding the agent’s functionality. One emerging idea is to integrate data from the Paris 2024 “Sport Coach” project to provide interactive and fun content to their members.	This innovation is not limited to process simplification alone, but also envisions a broader use of the agent, potentially making it accessible to the general public. The goal is to continue improving the structure and knowledge of resources, while maintaining a high level of performance.
The Importance of Support in Adopting the AI Agent	To ensure the successful adoption of this tool within the association, a support program is planned to assist members in their transition to this new tool. Despite the difficulty of assessing the time savings provided by the agent, initial user feedback indicates a significant improvement in information searches.	In short, the CNOSF demonstrates through this experiment that the integration of an AI agent is not only feasible, but also beneficial. Beyond process automation, a true digital transformation is taking place, promising a more efficient and user-centric future.