Many programming languages support different concurrency models. In practice these models are often combined, however the semantics of the combinations are not always well-defined. We studied the combination of futures and Software Transactional Memory. We introduce transactional tasks, a mechanism to create futures in a transaction. Transactional tasks allow the parallelism in a transaction to be exploited, while providing safe access to the state of their encapsulating transaction. We created ClojureTxTk, a fork of Clojure with support for transactional tasks. Furthermore, we ported two applications from the STAMP benchmark suite, and extended these to use transactional tasks: LabyrinthTxTk and BayesTxTk. Lastly, TxTkRedex is a machine-executable implementation of the operational semantics, in PLT Redex.