Great question.
There's a few approaches depending on your target latency. If you need to get below 2 seconds, yes, you'll likely want to use a WebRTC based technology to deliver the video. There's a couple of open source approaches out there, including Pion [1], which is a go implementation of a WebRTC stack. You could also build something on top of Jitsi [2]. Commercially there's also a few solutions, including Milicast [3], Red5Pro [4] and others.
The biggest problem with WebRTC based stacks is that the cross device compatibility is still generally poor, and the cost of operation is generally very high, as you can't use commodity CDNs for delivery.
If you're comfortable around the 2-5 seconds latency mark, there's more traditional HTTP based technologies available. MPEG-DASH has a Low Latency mode which uses chunk transferred HTTP fragments, and Apple has recently introduced a Low Latency HLS mode, which works in much the same way [5] [6]. You can build LL-HLS and DASH-LL solutions on top of open source toolchains like Streamline [7], and use commodity CDNs for delivery to reduce cost.
[1] https://github.com/pion
[2] https://jitsi.org/
[3] https://www.millicast.com/
[4] https://www.red5pro.com/
[5] https://mux.com/blog/low-latency-hls-part-2/
[6] https://mux.com/blog/the-low-latency-live-streaming-landscap...
[7] https://github.com/streamlinevideo/streamline
