This site tracks and sorts registered COVID-19 trials, harvested from multiple reliable global databases as well as other promising research not included in such sources, mined and pre-filtered using AI techniques. The studies are all manually reviewed by two humans for validity, reliability, methods, etc, and checked for duplication. They also regularize and standardize the language and data to make studies more easily and directly comparable. The result is then fed into an easily searchable online database. This provides what is essentially a super-fast and flexible way of conducting something akin to a systematic review (better, in some ways), and the results can be freely used by anyone interested in the current state of the research (good export facilities too). I rather like the idea that it is becoming a means for researchers themselves to connect with one another and coordinate research. For laypeople, it's a brilliant way to check the true state of research without the sensationalism or cherry picking of politicians or regular/social media. You can easily set filter and sorting conditions, and there are links to all the original data and papers (many of which are not paywalled). As of today, the site tracks 590 trials, but the number is growing all the time. The site and its features are still evolving, too. It has been built by researchers who did not have a wealth of web design expertise before this started, but you'd hardly know it: they have done a great job of getting it up and running and making it really usable and responsive.

You can read more about it in The Lancet.  I highly recommend the associated interview. The sound quality of the podcast is not great, but the interview is terrific, and it explains much more of the process, implications, and uses than the article itself. Some great reflections on the relative value of different kinds of data and many of the seldom stated complexities of scientific trials in general, including political and social issues, not to mention the immense promise of analytics approaches to greatly increase what we can learn from existing trials. Fascinating stuff.

Full disclosure: I am the very proud father of the second author of the paper and interviewee in the podcast, who is working punishing hours day and night to make this happen. In the fight against the pandemic he is only one among many heroes, but this one happens to be mine.