eBPF has significantly advanced networking and monitoring. However, its in-kernel execution, while stable and easy to deploy, limits performance compared to userspace alternatives like DPDK for network functions and userspace instrumentation for Uprobe. This session introduces bpftime, a userspace eBPF runtime designed for network functions and observability tools. bpftime allows existing eBPF control applications to operate in unprivileged userspace, offering Uprobe, syscall tracepoints, and XDP hooks with improved performance. It supports interprocess eBPF maps in userspace shared memory and can work with kernel eBPF maps, providing seamless integration and compatibility with the kernel's eBPF infrastructure.