@KrisianZurawski Hunting for the Lost City of Gold are you?
Datamine Discover for Pro - just get it if you use Pro to earn a living in a subsurface industry (geologist, etc). Cost a penny but if you are going to use it often it will pay for itself. Just make lots of backups as you go as it can be temperamental.
https://pages.dataminesoftware.com/discover-for-arc-gis-pro
You mention AHD - as in Australia? Datamine and the Discover team is local. I can connect you if you want.
Discover is good for typical resources projects but starts sucking when you want a cross section for a 300km rail line that doesn't have any drilling data.
Target for ArcGIS (Seequent) also an option but not my cup of tea.
Using Pro built-in tools it is possible but a pain to get an output of real geometry that you can then style & add to layouts for a typical mining cross section (that looks better than what Vulcan etc can do).
1 minute elevator method: Create 3d polylines that intersect everything you want to show sub-surface and update feature z using the RL. i.e. 1 for topo, 1 for base of weathering, 1 for each seam, 1 for each surface polygon such as geology, mining tenure (translated to topo). Then transpose the Y & Z dimension of each line. At this point you can apply a scale factor to Z and also set the section X origin to 0 and change the coordsys to something like web mercator (so Pro doesn't freak out).
The transpose can be done by converting the polylines to linear references, converting those to points, then flipping things if your sections are not straight or just a simple point along line to say every 25m and then flip them.
You now have a geometrically correct section dataset in basic ArcGIS 2d geometry that can be manipulated by all the Esri tools. Catch is that due to z-exaggeration being a hard-calculated value you have to do it all again if you decide to change your mind and change the scale. When I get a free & lucid moment I want to try to set it up as a Pro Notebook but that moment is elusive.
Non-vertical drillholes require them to be de-surveyed and you can transpose the origin to location on section, then go points to lines.
Then compile & style the data and manually draw polygon subsurface zones as required. PIA. I know but you get 'real' attributable datasets that you can use. Some of the Pro section views are slow and do funny things in layouts.
By attributing things properly you can then use map series to create a bunch of sections through a resource and add them to a layout with location map, vertical plan, horizontal section, dynamic legend, etc.
The fence diagrams and other Pro tools are handy but a lot of them create graphs or non-geometrical or non-attributable 'things' rather than geometry or become very hard to manage due to coordinate system issues.
If you have ArcMap use Xacto10 https://www.arcgis.com/home/item.html?id=54584a5e302e4014a495b8fc37fe0663#!
Useable but glitchy. Wish the person Open Sourced it as it has potential. Data just need a few attributes to be fit for automation once created.
In QGIS there is a lot of development going into OpenLog but current version is still a bit basic.
https://apeiron.technology/openlog/
Using an xlordb PostgreSQL database for the drilling data in OpenLog means you can use the data in Pro, QGIS, or ArcMap.
It may sound as if I have it all sorted but getting an efficient but robust section workflow in Pro that is suitable for good presentation output is a big hole in the Pro capability and I am not wanting to be a developer.
