IDFC FIRST Bank leadership talk: AI in banking beyond the buzzwords

The Leadership Talk with IDFC FIRST Bank, held at S.P. Jain Institute of Management and Research (SPJIMR) for management programme participants, gave them a very practical view of what AI in banking actually looks like, beyond the buzzwords. The session included Avik Sarkar, Chief Data and Analytics Officer; Rahul Dasgupta, Head – Credit Cards and Anti-Money Laundering (AML) Analytics; Anwesha Pal, Senior Vice President; Avantika Kapoor, alumnus and Product Manager (Gen-AI); and Sagar Nagrani, alumnus.

The session covered how the bank is using AI across different functions, from trade finance and current account (CA) onboarding to hiring, Credit Appraisal Memo (CAM) generation and enterprise data cleaning. The speakers noted that the bank does not treat AI as one large, standalone initiative. Different problems call for different solutions, including voice intelligence, document intelligence and knowledge bases.

The team also showcased three products. FIRST AI, an intelligent virtual banking assistant for customers, is built around the idea of ‘converse, not search’. Instead of simply helping customers find information, it can take actions, offer recommendations and speed up the overall customer journey. The ACE Document Agent uses optical character recognition (OCR) and form-filling capabilities to help banking teams process documents more efficiently, including in trade finance, where it scans, categorises and checks documents against a requirement list.

The speakers explained that AI is also making a measurable difference in credit and underwriting. AI-enabled underwriting has already helped the bank save around ₹30 crore. In corporate banking, automating CAM generation is helping relationship managers save significant time and spend more of it with customers. Meanwhile, the Credit Appraisal Department (CAD) uses AI to scrutinise credit structures. Importantly, the bank keeps humans in the loop, especially in document intelligence, so the models continue to improve.

A particularly interesting point raised during the session was that the bank also tracks First Time Not Right (FTNR) as a metric, showing that the focus is not just on automation but on getting processes right the first time.

One phrase from the session captured the approach well: ‘converse, not search’. It reflects a larger shift from AI simply providing information to AI actually helping people get things done.

Management programme participants found the session valuable for its grounding in real examples.

“What I liked most was seeing actual use cases rather than just hearing about AI in theory. It made the scale of AI’s impact on banking much more tangible,” said Chirag Singhla, PGDM 2026-28.

“The session showed that AI is not necessarily about replacing people. It is about taking away repetitive work so that people can focus on decisions, customers and higher-value tasks,” said Pranjal Banga, PGDM 2026-28.