Cisco tasks 27-year-old Paddy Krishnamurthy with AI-led code refactoring
Cisco has assigned 27-year-old Padmanabhan Krishnamurthy to refactor 98% of its legacy code within 24 months using unlimited AI tokens, a project that highlights a broader industry shift toward AI-driven modernization and new talent strategies.

Cisco has tasked 27-year-old Padmanabhan 'Paddy' Krishnamurthy with refactoring 98% of its legacy code using AI tokens within 24 months. Chief Product Officer Jeetu Patel gave Krishnamurthy unlimited AI tokens for the project and has four weeks to prove the concept with a two-person team.
Krishnamurthy joined Cisco in 2025 after the company acquired translation startup EzDubs, which he co-founded. He was hired from five levels below the executive rank. Patel seeks young hires who will challenge the system and the status quo. He believes experienced staff cannot ask questions that only someone with no experience would think to ask.
Team composition for code refactoring
The project combines three distinct talent groups to tackle the legacy code challenge. The team around Paddy consists of one-third experienced Cisco operators, one-third market systems thinkers, and one-third AI-native entrepreneurial founders and engineers. Patel believes that combining these three groups creates magic. Cisco aims to add more young, energetic hires to drive innovation and create constructive friction within teams.
Industry context on mainframe challenges
Financial institutions globally face massive legacy system burdens with significant staffing and documentation issues. Large financial institutions put 70% to 85% of their IT budgets toward maintaining older systems. 45 of the world’s 50 largest banks rely on mainframe applications for critical operations.
Douglas Santos, Itaú’s mainframe platform executive, said: "the bank had billions of lines of COBOL, a programming language from 1959, with scarce documentation." Systems written in COBOL still run an estimated 95% of U.S. ATM transactions. More than 40% of banks still run COBOL-based cores. 71% of mainframe teams report understaffing.
Proven AI modernization results
Tools like Batch2Cloud and AWS Transform have demonstrated dramatic efficiency gains in mainframe modernization projects. Itaú and GFT Technologies said their tool Batch2Cloud cut migration time for the bank’s batch systems by up to 70% and sped project delivery by up to 3.3 times. Application discovery that once took three months was done in two weeks using AWS Transform and Kiro, Amazon’s AI development tools. Business-rule extraction that required weeks of expert interviews now takes hours.
In applications where the tools were used end to end, migration time per application fell about 75%. Itaú said the generated code passed its security and quality pipeline on the first pass. The bank is on track to retire its mainframes by the end of 2028.
| Institution | Tool/Project | Result |
|---|---|---|
| Itaú | Batch2Cloud | Cut migration time by up to 70%, sped delivery 3.3x |
| Itaú | AWS Transform & Kiro | Cut 3-month application discovery to 2 weeks |
| General (Itaú) | AI tools end-to-end | Migration time per application fell ~75% |
| Danske Bank | AI agents | 60% overall acceleration, 4x faster forward engineering |
Talent strategy and generational dynamics
Cisco's leadership is actively reshaping team structures and recruiting approaches to drive innovation through generational friction. Patel creates tension between Gen Z and senior staff to generate friction on ideas. He treats every intern as a 'reverse mentor' and stated that Gen Z should be teaching veterans, not doing data entry.
This philosophy is echoed elsewhere. Josh Kushner, founder of Thrive Capital, hires people with less than four years of industry experience. Alon Chen, founder of Tastewise, hires entry-level Gen Zers with zero experience. Chen states that Gen Z has no boundaries or limitations in how they think about the world. Investor will.i.am believes the next big tech breakthroughs will come from young innovators at MIT and Stanford.
Internally, Cisco engineers using AI tools have reduced project build times by up to 20%. AI-assisted development has increased defect resolution throughput by 15 times. Approximately 85% of Cisco’s engineers were already using AI regularly as of earlier this year. The company expects 70% of its product portfolio to be fully AI-generated by the end of 2027. The project's budget line that changes is spending on AI tokens, not engineers. Cisco's deadline for completing 98% of its legacy code refactoring is 24 months.





