Acuity Analytics and Raiffeisen Bank on Moving AI from Experiments to Real Transformation

Jon O’Donnell from Acuity Analytics and Renato Rocha Souza from Raiffeisen Bank to discuss how banks are moving from early AI experiments to more meaningful transformation and what that really means in practice.

O’Donnell explains that most banks are already trialling multiple AI solutions, with several live across their organisations. Early use cases have focused on familiar, scalable work such as company profiles and research outputs – tasks that are well understood and easy to automate. These have been useful proof points, but leading banks are now pushing further into areas that genuinely differentiate them.

Many financial institutions began with AI assistants built on large language models and these tools have delivered solid returns by giving employees hands-on exposure to AI while automating everyday tasks. That combination of productivity and learning has been valuable, but O’Donnell notes the limitation: assistants tend to optimise single steps rather than transform whole process. As a result, banks are increasingly shifting towards workflow-based solutions that combine generative AI with automation and orchestration to redesign work end to end.

This is the focus of Acuity Analytics’s Agent Fleet, which O’Donnell says is already in production at organisations including Standard Chartered Bank and DC Advisory who are using AI agents to automate complete workflows such as credit risk assessments, investment memos and company profiles — unlocking far greater efficiency and impact than point solutions alone.

Souza adds the perspective of a bank actively implementing AI agents. Human oversight over AI is essential, especially as regulation evolves. While AI can take on more responsibility, accountability must always remain with people, highlighting a growing workforce challenge: junior, data-heavy roles are most at risk of automation, potentially creating a skills gap and increasing demand for senior expertise.

The discussion with Acuity Analytics and Raiffeisen Bank goes beyond efficiency into culture and ethics as Souza warns that AI models reflect the data, and values, they are trained on. Over-delegating judgement to AI risks locking organisations into a particular worldview shaped by the models they choose as different regional approaches to regulation and model development only heighten this concern.

Both Acuity Analytics and Raiffeisen Bank agree on one core principle: AI transformation must be people-led. Even in highly automated organisations, humans need to stay firmly in the driver’s seat, with AI as a co-pilot, supporting decisions, not silently defining the future.