How Custom AI Agents Fit into Today’s Teams

Forget ChatGPT – the next generation of finance is hinging on autonomous AI agents capable of complex tasks like auditing, algorithmic trading, FP&A, fraud detection, risk management, and regulatory compliance. Agentic AI is hardly an unfamiliar concept now; PwC’s AI agent survey (May 2025) of 300 senior executives found that 79% had already adopted AI agents into their systems, with 66% confirming that these tools delivered measurable value through increased productivity and cost savings. Gartner predicts that by 2028, 33% of enterprise software applications will include agentic AI, up from less than 1% in 2024.

At MAVI, we are seeing this evolution of finance first-hand. Our clients are not just hiring for capacity anymore; they are designing hybrid teams where skilled finance professionals gain efficiency in their decision-making by using an automated workflow. We’ve connected with Controllers who are using tools like Cursor to build niche agents to automate specific tasks such as report generation. These aren’t just one-size-fits-all products, and are instead custom tools that mold into unique business models, generating greater impact than run-of-the-mill AI. Here are a couple of examples of AI agents in finance in action:

• JPMorgan Chase uses “COIN” (Contract Intelligence) to automate contract review for credit risk and compliance, saving 360,000 hours annually, reducing legal costs and human-attributed mistakes, and accelerating approvals.
• LVMH deploys AI agents to adjust prices on its items based on factors like currency fluctuations to sustain profit margins during volatile market conditions.
• Mastercard employs generative AI-based predictive technology to protect 125 billion transactions annually against emerging threats, and has since doubled the detection rate of compromised cards, reduced false positives by up to 200%, and increased the speed of identifying merchants at-risk from fraudsters by 300%.
  

Still, adoption of custom AI agents has been slow and cautious in the broader industry. From conversations with finance leaders, MAVI has discovered that many are still very skeptical of available tools. The technology still isn’t quite there, and concerns surrounding cybersecurity, costs, and adoption have hampered adoption.

Regardless, custom AI agents are gradually taking on a growing mix of repeatable and strategic tasks, pushing companies to shift from larger functional departments to leaner, more tactical finance teams augmented by intelligent systems. But the introduction of tech does not replace humans; it simply redefines how they create impact.

At MAVI, we see accountants and analysts working on interpretation and higher-level decisions, while custom agents build the raw calculations at scale and provide data-driven insights in the background.

The future of finance will not be "man or machine" – it will be "man and machine" working in real-time, seamlessly together. If you're looking for someone to help build custom AI agents for your company's finance function, complete this form and be the first to access skilled professionals who can increase your team's efficiency with agents tailored to your needs.