HDFC Bank developed its proprietary AI platform, Neev, for an initial cost of ₹2 crore, bucking the trend of massive spending on large AI models. By focusing on specialized, domain-specific AI, the bank aims to enhance operational efficiency while maintaining strict control over banking data. This internal approach is designed to improve accuracy in critical tasks like trade documentation and customer service.
Detailed Coverage
HDFC Bank has shifted the narrative around artificial intelligence costs in the financial sector by revealing that its internal AI platform, Neev, was launched with an initial seed investment of only ₹2 crore. Ramesh Lakshminarayanan, the bank's Group Head of IT and Chief Information Officer, noted that ongoing maintenance and development costs have remained low. This strategy challenges the industry belief that banks must spend heavily on massive, general-purpose large language models to remain competitive.
Specialized AI Versus Large Models
The bank’s approach prioritizes building smaller, highly focused models tailored specifically for banking operations rather than relying on broad, expensive models. According to the bank, this design provides greater accuracy and lower processing delays, which are critical when managing complex tasks like verifying trade letters of credit or analyzing import and export documents. By developing Neev in-house, HDFC Bank maintains full ownership of its intelligence, allowing for better monitoring and compliance with evolving regulatory standards that demand high precision in financial services.
Operational Efficiency and Internal Ownership
Neev is built using a layered architecture that allows for reusable capabilities across different banking functions. The platform currently supports various tasks, including image extraction for documents like Aadhaar cards and assisting staff with complex credit card charge calculations. By automating these repetitive processes, the bank aims to manage higher volumes of business without a proportional increase in costs. The platform is managed by a small, dedicated team of 40 people, demonstrating how focused internal teams can deliver scalable technological solutions.
Workforce Strategy and Reskilling
As AI integration continues, the bank is focusing on reskilling its existing workforce rather than pursuing large-scale layoffs. Management has indicated that employees, including those with long tenures in specialized documentation roles, are being trained for new positions, such as customer-facing corporate advisory roles. The bank expects that natural attrition will manage headcount shifts while AI-driven productivity gains allow the institution to handle expanding business requirements. Investors will likely monitor whether these internal AI efficiencies translate into improved profit margins and reduced operational costs over the coming quarters. The bank's ability to maintain this cost-effective, in-house development model amid rapid sector-wide technology adoption remains a key area for long-term evaluation.