Real challenge remains in the implementation of accurate and efficient methods for extracting information from historical handwritten tabular records. We propose a human-in-the-loop workflow with benchmarks of open- and closed-source methods for extracting and processing handwritten tabular data.