Building Quantum Workflows using Composable Software within the Qiskit Patterns Framework

Abstract

Delivering useful quantum computing requires turning key capabilities into software and effectively stitching them together to explore problems at scales that are hard classically. In this tutorial, we will introduce Qiskit Patterns as a framework for breaking down utility-scale quantum workloads and contextualizing the necessary capabilities into four main steps – mapping domain-specific problems to quantum circuits and operators, optimizing the circuits for the target quantum hardware, executing, and post-processing the results. We will discuss the need for developing software building blocks that address tasks within these steps and that can be composed together to build larger workflows. Then, we will highlight several software building blocks that address these steps, examples of which may include building quantum circuits with the Qiskit circuit library, transpiling circuits through the latest AI and heuristic methods, executing on quantum hardware via the Qiskit Runtime Primitives, applying error mitigation techniques such as TREX, ZNE, and PEC, and using circuit knitting techniques to further optimize circuits for execution. Participants will be introduced to the main concepts behind each of these building blocks, and will be shown how to access and configure them in Qiskit. Finally, we will show through concrete examples how such building blocks can be composed together to construct larger workflows.