About one month to get an MVP running and start testing. We then support ongoing additions of more data, features, and tuning from your in-house data science and ML team.