In the era of the Internet of Things (IoT), data from sensors can give insightful enterprise information through analytics. As a result, several enterprises are adopting sensors and other wireless technologies for their needs. However, some challenges exist within the IoT space. Most of the devices in use have varied device semantics and protocol variations which can limit interoperability. As a result, data and process integration can be hindered. In this work, we propose a middleware with both machine-to-infrastructure (M2I) and machine-to-machine (M2M) capabilities which addresses these problems based on mapping techniques between the heterogeneous device semantics and providing a common interface for merging protocol variations. When a device is discoverable, our middleware uses an enhanced environment-context features to match the appropriate communication protocol. This aids in the pushing of data from within-range sensors to a cloud-hosted infrastructure.