Deutsche Bank is experiencing a remarkable transformation through artificial intelligence. Work that previously took two years can now be completed in just three to six months. Artificial intelligence has boosted productivity at Deutsche Bank, allowing tasks that once took years to be completed within months, a senior executive said Thursday.
German financial institutions are leveraging AI to accelerate technology projects and tackle internal backlogs, but remain wary of rising computing costs.
“What used to take two years is now being done in three to six months. We know it’s more productive,” said Dennis Lu, chief information officer for investment banking at Deutsche Bank, speaking on the sidelines of the bank’s Bank on Tech event in Bangalore, India. He declined to quantify the impact.
Backlogs that once took months can now be cleared in weeks, Lu said, adding: “All I want is to continue to use these tools to streamline things.”
The bank has approximately 9,000 employees in the Indian technology sector, accounting for approximately 45% of the global technology sector workforce. Global companies are increasingly turning to Indian hubs for higher-value functions such as finance, software development, and research and development.
Still, managing AI deployment costs is a priority as providers move to pay-as-you-go pricing models, Lu said, likening it to the discipline businesses developed during their transition to cloud computing.
AI companies like Anthropic and OpenAI are moving toward token-based pricing, where they charge customers based on usage rather than subscription-based services.
Deutsche Bank engineers are allocated token quotas and can request additional capacity, but they must demonstrate value and what they learn will then be shared across the organization, Lu said.
“We’re monitoring usage patterns… We don’t want to slow people down, we want them to continue, but we also want to benefit them,” he said.
The bank is also developing AI tools to automate tasks such as financial data extraction and analysis, as well as applications that link external events such as geopolitics and market trends to portfolios to understand exposures.
Lu said banks will remain cautious about introducing AI into everything, using simpler models for day-to-day operations and also evaluating where traditional solutions could be more effective.
