In vitro methods for developing binding domains have been well-established for many years, owing to the cost-efficient synthesis of DNA and high-throughput selection and screening technologies. However, generating high-affinity binding domains often requires the development of focused maturation libraries for a second selection, which typically demands a detailed understanding of the binding surfaces from the initial selection, a process that can be time-consuming. In this study, we accelerated this process by using deep sequencing data from the first selection to guide the design of the maturation library. Additionally, we employed a high-throughput screening system using flow cytometry based on Escherichia coli display to identify conditional binding domains from the selection output. This approach enabled the development of a high-affinity binder targeting the cancer biomarker HER3, with a binding affinity of 3.3 nM at extracellular pH 7.4, 100 times higher than the first-generation binding domain. Notably, the binding domain features a pH-dependent release mechanism, enabling rapid release in slightly acidic environments (pH ≈6), which resemble endosomal conditions. When conjugated to the cytotoxin mertansine (DM1), the binding domain demonstrated specific cytotoxic activity against HER3-expressing cell lines, with an IC50 of 2–5 nM. The presented approach enables the efficient development of conditional binding domains which hold promise for therapeutic applications.