Cisco President and Chief Product Officer Jeetu Patel backed a proposal to use AI and a two-person team to refactor 98% of Cisco’s legacy code within 24 months, Fortune reported.
The proposal came from Padmanabhan Krishnamurthy, a former startup co-founder. Fortune said Patel told it he had given Krishnamurthy “unlimited tokens” for the work. The publication reported that Krishnamurthy believed a two-person team could rewrite the code. The 98% figure is a proposal target, not a completed result. Fortune did not identify which code counted toward the percentage or how Cisco would measure progress.
Fortune described Krishnamurthy as 27 in its Oct. 30 story and reported that he co-founded EzDubs in 2023. It said he joined Cisco after the company’s acquisition. Cisco announced it completed the deal on Nov. 25, 2025, describing EzDubs as a San Francisco-based AI software company with a speech-to-speech product that translated calls, voice messages and texts in real time. Cisco said the acquisition would expand its collaboration portfolio with product technology and engineering talent. Its public announcement did not mention the code project.
Earlier, Krishnamurthy and fellow undergraduate Amrutavarsh Sanganabasappa Kinagi developed “Helen,” an AI lip-reading device, Hong Kong University of Science and Technology (HKUST) said. The university said the project won its 2019 President’s Cup. A camera sent video to a separate device, where lip-reading software converted it into text.
A December 2022 master’s thesis by Krishnamurthy, listed on the CV of Columbia University computer science professor Lydia B. Chilton, focused on multilingual text-to-speech synthesis and AI dubbing.
Patel told Fortune that early-career employees can challenge established practices, while different viewpoints can create tension within teams. He said, “Tension is good.” Fortune also reported a four-week period to demonstrate the approach but gave no start date or outcome.
