The slowest step in natural history collections digitization is converting imaged labels into digital text. This long-recognized efficiency bottleneck can be overcome and we present a working solution that leverages synergies between citizen science efforts and machine learning approaches. We present two new semi-automated services. The first detects and classifies typewritten, handwritten or mixed labels from herbarium sheets. The second uses a workflow tuned for specimen labels to OCR label text. The label finder and classifier was built via humans-in-the-loop processes that utilize the citizen science Notes from Nature (NFN) platform to develop training and validation datasets to feed into a machine learning pipeline. Our results showcase >93% success for finding and classifying main labels. The OCR pipeline optimizes pre-processing, multiple OCR engines and post-processing steps, including an alignment approach borrowed from molecular systematics. This pipeline yields >4-fold reductions in errors compared to off-the-shelf open source solutions.  The OCR workflow also has a human validation using a custom NFN tool. Our work showcases a usable set of tools for herbarium digitization including a custom-built web application accessible to all. Further work to better integrate these services into existing toolkits can support broadest community use.