Two years ago we made a decision in Salesforce to build an entirely new optimization engine for our Field Service solution. It took about 20 years to build the current engine which includes very rich functionality and solves many use cases. It was clear that reaching the same level of functionality will require a substantial investment of time and effort and so we found ourselves in a difficult conundrum - How do we swap a great and well-accepted engine with a new unfamiliar one, in minimum time and still provide added value to our customers?