The paper presents a novel technique, MEPS@NLO, for combining next-to-leading order (NLO) parton-level calculations with parton showers in hadron-hadron collisions. This method avoids double counting by using a modified truncated shower scheme, ensuring both fixed-order accuracy and logarithmic accuracy of the parton shower. The approach is applied to $W$-boson production at the Large Hadron Collider (LHC) using the SHERPA event generator. The results show a significant reduction in theoretical uncertainties compared to existing methods, highlighting the predictive power of the new technique. The paper also discusses the renormalization and factorization scale dependence of the approach and provides a detailed implementation in SHERPA, demonstrating its effectiveness in describing multi-jet events with high precision.The paper presents a novel technique, MEPS@NLO, for combining next-to-leading order (NLO) parton-level calculations with parton showers in hadron-hadron collisions. This method avoids double counting by using a modified truncated shower scheme, ensuring both fixed-order accuracy and logarithmic accuracy of the parton shower. The approach is applied to $W$-boson production at the Large Hadron Collider (LHC) using the SHERPA event generator. The results show a significant reduction in theoretical uncertainties compared to existing methods, highlighting the predictive power of the new technique. The paper also discusses the renormalization and factorization scale dependence of the approach and provides a detailed implementation in SHERPA, demonstrating its effectiveness in describing multi-jet events with high precision.