The AI in Business Podcast From Manual Reports to Generative and Agentic AI Automation in Finance - with Pavlé Sabic of Moody’s
13 snips
Sep 11, 2025 Pavlé Sabic, Senior Director of Generative AI Solutions at Moody's, dives into how agentic AI is transformation financial workflows. He discusses how large enterprises tackle issues like fragmented data and manual inefficiencies through AI automation. Pavlé highlights impressive tools for credit memo automation that cut production time by 60%, and techniques for early risk detection in portfolio monitoring. He emphasizes the strategic edge of proprietary data in finance and the critical need for human oversight in AI implementation.
AI Snips
Chapters
Transcript
Episode notes
Enterprise Pain Points Drive Agentic AI
- Financial firms face manual inefficiencies, fragmented data, frequent regulatory change, and analyst overload that create risk and slow decisions.
- Agentic AI combined with intelligent documentation addresses scale, accuracy, and real-time strategic decisioning.
Design For Auditability First
- Build agentic systems with auditable workflows so probabilistic LLM outputs become consistent and traceable.
- Prioritize governance and stepwise auditing to make LLM-driven outputs acceptable for financial compliance.
Credit Memo Workflow Cuts Production Time
- Moody's implemented a human-led agentic workflow for a large bank that synthesizes origination documents, industry trends, and news.
- Analysts supervise the output and production time fell by about 60% while humans retained final judgment.