Hey all! I'm doing some research on boosting developer productivity and simplifying the onboarding process for new projects. I'm looking into an idea where we can streamline the setup that usually takes days due to outdated documentation and missing dependencies. The idea is to provide a Git URL and have an AI scan the codebase for manifests, ports, and database strings. It would then infer the architecture and generate all the necessary Dockerfiles along with a fully linked docker-compose.yml. This would allow someone to go from cloning a repository to running a local simulation in minutes without any manual configuration. I'm curious: is there currently a tool that covers this entire process from repo to orchestration? What technical challenges do you think could be deal-breakers for an AI-generated setup? If such a tool existed and was reliable, would you find it useful for onboarding developers? Thanks for your insights!
4 Answers
Claude might be able to help with this! I had it set up an MSSQL server and it handled the project build in another container while deploying everything simultaneously. It worked pretty smoothly!
I haven't seen any specific AI designed for this. If you try out Claude Code, just keep in mind that AIs can be unpredictable. I’d set everything up once and save those Dockerfiles and the docker-compose.yml so you can tweak them later. It’s much easier than dealing with 'it worked on my machine' issues. Do you really think this kind of setup might have issues if you don't save those files?
While not strictly AI, have you thought about using Ansible? It can automate the setup process quite effectively.
Have you considered using a devcontainer instead? It seems like a better way to start without all the local setup hassle.
Thanks for your insights! That makes sense—having saved configs could really help avoid those 'it worked on my machine' problems!