That's basically what I do with my own tool. I create my own data files and store lots of meta information like similar words, usage, and mnemonics in there. I try to keep that somewhat usable for others, but ultimately, I am building the tool for myself and also write the data about words in the way I personally find it useful, so it will be biased in that way. Only, that I am also building the software to use the data around it, combining 2 hobbies, computer programming and learning Mandarin.