Complementing Chapter 16, this chapter introduces algorithms that an individual buyer can utilize to manage a budget constraint over a sequence of auctions. These are also the types of algorithms offered by internet platforms for bidding on behalf of buyers. We introduce the dual mirror descent framework, and show that it leads to no-regret guarantees when the incoming auctions are generated from a fixed stochastic model, and competitive ratio guarantees under adversarial input. We discuss extensions to return on spend constraints and predictions about traffic flow, such as the spend plans utilized in the ad auction industry.