Artificial intelligence is doing more than helping developers write code faster. At Deutsche Bank, it is transforming the pace of enterprise software delivery, allowing technology initiatives that previously required years to complete to move into production within just a few months.

Speaking at the Bank on Tech event in Bengaluru, Denis Roux, Chief Information Officer for the Investment Bank at Deutsche Bank, explained that AI is helping the organization accelerate technology projects, reduce engineering backlogs, and improve overall productivity. According to Roux, projects that once took around two years can now be delivered in as little as three to six months, while development queues that previously stretched across several months are being resolved within weeks.

The results highlight a broader shift taking place across enterprise technology. AI is becoming a practical engineering tool that supports software delivery throughout the development lifecycle, from implementation and testing to documentation and operational improvements. Rather than replacing software engineers, these systems allow development teams to focus more of their time on solving complex business problems while reducing repetitive manual work.

Productivity Gains Still Require Cost Control

While AI is delivering measurable efficiency improvements, Deutsche Bank is treating adoption with the same financial discipline that organizations applied during their transition to cloud computing.

As leading AI providers increasingly move toward token-based pricing models, where usage directly determines cost, the bank closely monitors how engineers consume AI resources. Developers receive allocated token budgets and can request additional capacity when they demonstrate clear business value. Successful use cases are then shared across the organization to encourage efficient adoption without unnecessary spending.

This reflects a growing challenge for enterprise AI programs. Faster software development can generate significant returns, but only when organizations actively manage infrastructure costs, usage patterns, and governance.

Choosing the Right AI for the Right Job

Deutsche Bank is also expanding AI into operational workflows beyond software engineering. The organization is building AI-powered solutions that automate financial data extraction and analysis while developing systems capable of connecting external events, including geopolitical developments and market movements, with portfolio data to improve exposure analysis.

At the same time, the bank is avoiding a one-size-fits-all approach. Simpler AI models are used for routine tasks, while traditional software continues to handle processes where conventional solutions remain the better choice. This balanced strategy helps ensure that AI is deployed where it creates measurable value rather than simply following industry trends.

What This Means for Enterprise Software

The experience at Deutsche Bank reinforces an important lesson for organizations investing in AI.

The biggest productivity improvements rarely come from adding AI to isolated tasks. They come from integrating AI into mature software engineering practices supported by clear governance, cost visibility, and strong delivery processes.

As AI capabilities continue to improve, organizations that combine engineering discipline with thoughtful AI adoption are likely to reduce delivery timelines, improve operational efficiency, and build software that scales more effectively over the long term.

Source