Key Takeaways: From Pilot to Project - How Financial Leaders Drive AI Impact, Not Abandoned Projects
In our webinar, From Pilot to Project: How Financial Leaders Drive AI Impact, Not Abandoned Projects, hosted on September 10th, Oliver Munn, Director of AI Capability at Faculty, shed light on a critical problem facing the industry: the high rate of AI project failure. The high-profile July 2025 MIT study, which revealed that 95% of organisations receive zero return on their AI investments, is but one of many indicators that companies have so far struggled to bank the benefits of AI. Oliver presented a compelling case for a different approach focused on Agentic AI.
The Agentic AI Paradigm
Oliver introduced the concept of an AI agent as an intelligent system that can independently accomplish tasks on your behalf. These are not just sophisticated chatbots, but powerful Large Language Models (LLMs) equipped with tools like APIs and search capabilities, enabling them to take action. Early adopters are already seeing significant returns, with companies like Klarna and J.P. Morgan leveraging these systems for a variety of tasks, including customer service, coding, and information processing.
A Real-World Application
To demonstrate the power of this approach, the webinar featured a fascinating demo of an Agent Command Centre in an insurance workflow. The system used a suite of specialised agents to automate the handling of broker submissions. For instance, a Pre-Underwriting agent automatically processes incoming emails, extracts and enriches data, and even communicates with brokers to gather missing information. A Quote Generation agent then takes over, preparing a quote for a human underwriter’s review and approval. This hybrid model allows for high-value automation while maintaining crucial human oversight on risky, high-stakes tasks.
Agentic Success
Successfully integrating AI agents isn't just about the technology; it’s about a strategic framework. Oliver likened the process to bringing on a new employee, highlighting three essential factors for success. First, context is paramount; the agent must be trained on your specific business rules, policies, and regulations to operate effectively. Second, a clear job definition is required, outlining the agent's precise scope, responsibilities, and level of autonomy. Finally, a robust feedback loop is vital for measuring the agent's performance, continuously improving its rules, and ensuring it consistently performs the correct actions.
Transforming AI from Projects to Performance
The webinar makes a case that the future of AI in finance lies in moving beyond isolated pilot projects and embracing a strategic, agentic approach. By treating AI as a capable, accountable partner rather than a simple tool, financial leaders can drive real business value, steer genuine transformation, and ensure their investments yield meaningful, quantifiable returns. You can watch the full webinar recording here.
Our Artificial Intelligence for Leaders course provides you with the strategic knowledge and practical frameworks to successfully implement AI, build effective teams, and drive real business impact. Learn how to transform your business with AI, moving from abandoned projects to impactful performance.
