The emergence of SARS-CoV-1 in 2002 and SARS-CoV-2 in 2019, alongside evolving viral variants, underscores the ongoing threat posed by novel coronaviruses. Monoclonal antibodies have proven effective against these viruses, but most are derived from convalescent patients' B cells, which represent a limited resource during the early stages of an outbreak. This highlights the need for a robust platform to swiftly discover potent virus-neutralizing antibodies. We propose a strategy to leverage pre-existing neutralizing antibodies targeting viruses that utilize the same receptor entry mechanism. Although three anti-SARS-CoV-1 antibodies (80R, m396, and S230) did not bind to the SARS-CoV-2 RBD, we employed bacterial display-based high-throughput directed evolution and successfully isolated two antibodies, IJ4G and IJ225, which exhibited a strong binding affinity to both SARS-CoV-2 wild-type and the Delta variant, demonstrating potent neutralization activity by effectively disrupting ACE2 binding. Notably, these antibodies also retained high binding affinity for the SARS-CoV-1 RBD. Additionally, structural docking models revealed specific amino acid interactions that enabled these engineered antibodies to bind to both SARS-CoV-1 and SARS-CoV-2 RBDs, adapting to the sequence differences between the two. Our findings demonstrate the feasibility of rapidly developing potent neutralizing antibodies using only the virus spike protein sequence during the early stages of virus emergence, offering a promising approach for pandemic preparedness.