This paper presents methods for suppressing correlated coherent errors in quantum computers using context-aware compiling. Coherent errors, especially those that occur in correlation among qubits, are detrimental to large-scale quantum computing. The authors experimentally characterize various error sources and theoretically connect them to the physics of superconducting qubits and gate operations. Based on this knowledge, they develop compiler strategies to suppress these errors using dynamical decoupling or error compensation, taking into account the context of each layer of computation, such as qubit connections, crosstalk terms, and gate or idle periods. The authors introduce two compiler strategies: Context-Aware Dynamical Decoupling (CA-DD) and Context-Aware Error Compensation (CA-EC). CA-DD suppresses coherent errors by applying dynamical decoupling sequences tailored to the circuit context, while CA-EC compensates for errors by modifying existing gates to absorb their inverse. These strategies are effective in reducing error rates, as demonstrated by a 18.5% increase in layer fidelity for a 10-qubit circuit layer. The paper also discusses the challenges of suppressing coherent errors in quantum circuits, including the need to avoid applying DD gates to qubits that are already actively participating in a gate, and the potential for DD sequences to introduce systematic errors. The authors show that CA-EC is particularly effective for errors that are well-characterized and remain constant over time, such as always-on ZZ and Stark shift errors. The authors apply their compilation strategies to several quantum applications, including the simulation of 1-D Ising chains and Heisenberg rings, the estimation of circuit layer fidelity, and circuits with intermediate measurements and feedforward. In all cases, they demonstrate a reduction in errors and a decrease in error mitigation overhead. The paper also discusses the experimental methodology used to characterize and suppress errors, including the use of the IBM Quantum Platform and the characterization of various error sources such as AC Stark shift, next-nearest neighbor ZZ interaction, and slow Z oscillations from charge-parity fluctuations. The authors conclude that context-aware compiling is a promising approach for suppressing correlated coherent errors in quantum circuits, significantly reducing the overhead of error mitigation and correction. The methods presented are applicable to a wide range of quantum applications and can be integrated into existing error mitigation protocols.This paper presents methods for suppressing correlated coherent errors in quantum computers using context-aware compiling. Coherent errors, especially those that occur in correlation among qubits, are detrimental to large-scale quantum computing. The authors experimentally characterize various error sources and theoretically connect them to the physics of superconducting qubits and gate operations. Based on this knowledge, they develop compiler strategies to suppress these errors using dynamical decoupling or error compensation, taking into account the context of each layer of computation, such as qubit connections, crosstalk terms, and gate or idle periods. The authors introduce two compiler strategies: Context-Aware Dynamical Decoupling (CA-DD) and Context-Aware Error Compensation (CA-EC). CA-DD suppresses coherent errors by applying dynamical decoupling sequences tailored to the circuit context, while CA-EC compensates for errors by modifying existing gates to absorb their inverse. These strategies are effective in reducing error rates, as demonstrated by a 18.5% increase in layer fidelity for a 10-qubit circuit layer. The paper also discusses the challenges of suppressing coherent errors in quantum circuits, including the need to avoid applying DD gates to qubits that are already actively participating in a gate, and the potential for DD sequences to introduce systematic errors. The authors show that CA-EC is particularly effective for errors that are well-characterized and remain constant over time, such as always-on ZZ and Stark shift errors. The authors apply their compilation strategies to several quantum applications, including the simulation of 1-D Ising chains and Heisenberg rings, the estimation of circuit layer fidelity, and circuits with intermediate measurements and feedforward. In all cases, they demonstrate a reduction in errors and a decrease in error mitigation overhead. The paper also discusses the experimental methodology used to characterize and suppress errors, including the use of the IBM Quantum Platform and the characterization of various error sources such as AC Stark shift, next-nearest neighbor ZZ interaction, and slow Z oscillations from charge-parity fluctuations. The authors conclude that context-aware compiling is a promising approach for suppressing correlated coherent errors in quantum circuits, significantly reducing the overhead of error mitigation and correction. The methods presented are applicable to a wide range of quantum applications and can be integrated into existing error mitigation protocols.