Apache Spark Workflow for Real-time Analytics

I'm looking for an expert in Apache Spark to help me create a base workflow execution library. This library is primarily aimed at managing workflows which inherently run Apache spark pipeline

Key Requirements:
- Extensive experience with Apache Spark
- Strong understanding of real-time data processing
- Ability to design a flexible and reusable workflow execution library

Although specific machine learning tasks and datasets have not been defined yet, the library should be designed with potential support for:
- Classification
- Regression
- Clustering

And interaction with:
- SQL databases
- NoSQL databases
- CSV files

Please include in your proposal any relevant experience you have, as well as examples of similar projects you've completed.