High Speed Two (HS2) is the new high-speed railway line in the UK currently being constructed between London and Birmingham. The designs of many HS2 retaining structures, as is the case for many projects in the UK, were progressed concurrently with ground investigation. In this situation, to manage uncertainties in the ground, a cautious assessment of ground parameters was usually adopted. The purpose of field monitoring varies depending on the responsible or interested party. For designers, the primary aim is usually for design verification of their permanent works design. This paper describes how field observations, particularly quality instrumentation and monitoring data, were used at HS2 Bromford Tunnel East Portal to improve the existing design as construction progressed to create savings. To improve the efficiency in data processing, DAARWIN, a machine learning-based technology was used. The design and construction of the eastern end of the Bromford Tunnel East Portal were modified by eliminating the requirements of temporary support using measured field data, site observations and a carefully controlled construction through the application of the Observational Method. The portal construction programme was shortened by at least two weeks despite only three of the remaining temporary steel props being omitted. The findings revealed opportunities to extend the application of the Observational Method on adjacent structures including Bromford Tunnel West Portal and Washwood Heath Retained Cut. Given the great length of retained excavations of the two structures, the potential efficiency gains in cost, time and carbon emissions are significant whilst enhancing safety.