This is what a successful compound looks like in the clinic: https://www.ncbi.nlm.nih.gov/pubmed/3299089

The patient group was people with AIDS. 140 people received azidothymidine, 137 in the control group. 19 people in the control group died, 1 in the group that received AZT. Meanwhile it's not quite clear if the 8 % death in the remdesivir group is an improvement over the 11 % of deaths in the placebo group. Doesn't matter, as long as Gilead shareholders get a boost. It's so bad that it isn't funny.

"remdesivir’s polymerase inhibitor mechanism can reduce viral load, experts said. However, all available data is still empirical in nature, and load reduction may not translate into clinically relevant outcomes like reducing hospital stay or decreasing mortality (https://www.pharmaceutical-technology.com/comment/gilead-rem...)

Remdesivir is an element similar to one the cells use to create protein chains. When used instead of the real element it stops the chain. This limits the ability of the virus to create new copies. It would be like putting a cinderblock in a brick wall.

Unfortunately, SARS-Cov-2 has a "spell checker" that can remove the "mis-spelled" element (aka the cinderblock) and continue building the chain.

Remdesivir is difficult to make. About 8 liters of starting material is used to create a milli-liter of product which is about 1 injections. It appears that each patient may require multiple injections.

Remdesivir is not a vaccine. A Vaccine "trains" your immune system to combat the virus. Remdesivir is an injection and likely has to be used multiple times.

Remdesivir might reduce viral load (number of virus particles). The relationship between viral load, which is high in the beginning but less later, does not necessarily translate to sickness or death. Remdesivir might be useful within 48 hours of exposure but symptioms don't shou up for 5 days. Peak viral load has passed by the time symptoms occur.