taosdata/tdengine

TDengine is an open-source database purpose-built for storing and querying time-series data, data that arrives as a continuous stream of readings stamped with a timestamp, like sensor measurements, vehicle telemetry, or stock prices. Traditional databases struggle when billions of devices are each sending readings every second, causing slowdowns and massive storage costs. TDengine is engineered to handle that volume efficiently, claiming to outperform other time-series databases on ingestion speed, query performance, and compression. It is cloud-native, meaning it was designed from the ground up to run across multiple servers (distributed) and can be deployed on Kubernetes or public cloud platforms. A built-in AI agent called TDgpt adds forecasting and anomaly detection directly inside the database, you query for patterns or predictions the same way you query for raw data. Other built-in features include caching, streaming data processing, and data subscription (similar to how apps listen for real-time updates). You would use TDengine if you are building an industrial IoT platform, a connected-vehicle monitoring system, or any application where millions of sensors are generating time-stamped readings and you need to query and analyze that data in real time without a sprawling stack of separate tools. It runs on Linux, macOS, and Windows, and is written primarily in C.