It sounds like the person asking this question is about as junior as the team is (in choice of words, ideas about rewriting and best practices). It's hard to gi

It sounds like the person asking this question is about as junior as the team is (in choice of words, ideas about rewriting and best practices). It's hard to give specific advice without knowing more about the situation, for example what role the questioner has in the team. It is a tech lead, a boss, or a project lead? So let's stick to advice that is universal to all roles and I think most people who have been in similar situations would agree with. First, let's be clear about one thing: This situation isn't the least bit unusual. From the facts above it doesn't look very bad. The team is small, and you can all gather in the same room and communicate. The fact that there is no framework and no patterns in place is good given the circumstances, awful codebases based on ancient frameworks and legacy patterns are generally an order of magnitude more work to understand. Second, be humble towards the team and the problem. After such a long time, there's bound to be details that you don't know, and you have to find out about them sooner rather than later. People may seem resistant to change, but understand their angle and work with them. The likely want their codebase to improve, too, even if they see other problems as more pressing. It all depends on what your role is, and if you intent to help out with the actual work or not. But again, this is a small team with a shared goal. Third, start with the lowest hanging fruit. Personal opinions come into play here, but I probably would look at operational issues early. Get monitoring in place. Test backups (yes, really). Some key metrics, both application wise (on some key processes such as login or payments) and operational (memory, open files, sockets). Learn about version control and start using it. Get proper test environments in place (including databases and mocked external integrations). Good luck! Things are probably not as bad as you think. This type of work is really quite rewarding, because results are quickly very visible to everyone.