Repo here https://github.com/corpulent/fragments
The project is a tool intended to feed BI apps and dashboards, but I am exploring other domains. It was born out of a solution I built for a client, I answer alot of questions in this reddit post if anyone is interested https://www.reddit.com/r/dataengineering/comments/v1xjub/sharing_my_data_engineering_projectidea_using/.
Users can connect various data sources, like databases, third-party apps, documents, s3 and so on, write some transformation code and deploy. Everything is encapsulated "optionally" in a jinja style format so you dont need to be a data engineer to use this. These jinja models can be very simple like {"hello": "world"} to very complex with custom filters/logic, multiple data sources, and DAGs.
The project is in its early stage, and I am just collecting thoughts and feedback. Let me know you guys think.