With the explosion of biological data, the primary challenge is not how to store the data, nor what computational resources to use to process them, but it is the general lack of researchers' understanding to manipulate and analyse these data. This problem could be solved with comprehensive and up-to-date bioinformatics training. How can we build such an infrastructure? Over the last few years, I had the opportunity to lead and contribute to several training programs and communities in bioinformatics. In this talk, I will share some lessons learned and ideas to build usable, comprehensive, and empowering bioinformatics training infrastructure. I will talk about community-driven development of free, FAIR and reusable material, online and hybrid training, mentoring, and further illustrate these concepts with concrete examples from my work.