https://www.jimmunol.org/content/early/2020/09/03/jimmunol.2...
As with any pandemic there will be a small minority of outliers who experience symptomatic reinfections.
The evidence is actually scant but there's a decent discussion of it in the Guardian article I linked below and the paper linked from it.
This novel coronavirus is different enough from others, and the human body's response to it is very unusual, so it's reasonable to ask whether infection confers immunity.
People were quick to assume immunity at first, because that's what people want to be true. But it was recognised as a dangerous assumption, which has a realistic prospect of being wrong.
Some people end up with antibody levels too low to measure after infection, which adds more weight to the idea that they might not be immune afterwards. You could think of that as a level of evidence; certainly it comes under "we should take this seriously". (Fortunately the human body has other mechanisms for long-term immunity (T-cell memory), but that still needs to be studied.)
The difference between assumption and evidence on this is important because it strongly informs our models of its propagation, how many people will get it how fast, how best to contain and limit it, and whether people who have had can safely (for themselves and others) behave differently afterwards.
If you've had Covid-19 and you're still struggling with Long Covid 3-6 months later, you're going to want to know if there's a realistic chance you could catch it a second time, especially because by then you know you are one of those susceptible to awful symptoms. It's not enough to assume it's just like other coronaviruses - you will have personal experience that it isn't.
https://www.theguardian.com/commentisfree/2020/aug/27/corona...
With widespread serologic testing, we can at least get some waypoint data, and for people like you some peace of mind.